Index

NUMERICAL MODELING AND PERFORMANCE ANALYSIS
OF SOLAR-POWERED IDEAL ADSORPTION COOLING SYSTEMS

A THESIS SUBMITTED TO
THE GRADUATE SCHOOL OF NATURAL AND APPLIED SCIENCES
OF
MIDDLE EAST TECHNICAL UNIVERSITY

BY

ONUR TAYLAN

IN PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR
THE DEGREE OF MASTER OF SCIENCE
IN
MECHANICAL ENGINEERING

MAY 2010

Approval of the thesis:


submitted by ONUR TAYLAN in partial fulfillment of the requirements for the
degree of Master of Science in Mechanical Engineering Department, Middle
East Technical University by,

Prof. Dr. Canan zgen ___________
Dean, Graduate School of Natural and Applied Sciences

Prof. Dr. Suha Oral ___________
Head of Department, Mechanical Engineering

Assoc. Prof. Dr. Derek K. Baker ___________
Supervisor, Mechanical Engineering Dept., METU

Prof. Dr. Bilgin Kaftanolu ___________
Co-supervisor, Manufacturing Engineering Dept., Atlm University

Examining Committee Members:

Assist. Prof. Dr. lker Tar _______________
Mechanical Engineering Dept., METU

Assoc. Prof. Dr. Derek K. Baker _______________

Prof. Dr. Bilgin Kaftanolu _______________
Manufacturing Engineering Dept., Atlm University

Assist. Prof. Dr. Ahmet Yozgatlgil _______________

Assist. Prof. Dr. Eray Uzgren _______________
Mechanical Engineering Dept., METU Northern Cyprus Campus

Date: _______________
iii

I hereby declare that all information in this document has been obtained and
presented in accordance with academic rules and ethical conduct. I also declare
that, as required by these rules and conduct, I have fully cited and referenced
all material and results that are not original to this work.

Name, Last name : Onur TAYLAN

Signature :
iv
ABSTRACT

Taylan, Onur
M.S., Department of Mechanical Engineering
Supervisor : Assoc. Prof. Dr. Derek K. Baker
Co-Supervisor : Prof. Dr. Bilgin Kaftanolu

May 2010, 139 pages

Energy consumption is continuously increasing around the world and this situation
yields research to find sustainable energy solutions. Demand for cooling is one of the
reasons of increasing energy demand. This research is focused on one of the sustain-
able ways to decrease energy demand for cooling which is the solar-powered adsorp-
tion cooling system. In this study, general theoretical performance trends of a solar-
powered adsorption cooling system are investigated using TRNSYS and MATLAB.
Effects of different cycle enhancements, working pairs, operating and design condi-
tions on the performance are analyzed through a series of steady and seasonal-
transient simulations. Additionally, a normalized model is presented to investigate
the effects of size of the system, need for backup power, collector area and mass of
adsorbent. Results are presented in terms of values and ratios of cooling capacity
weighted COP. For the conditions explored, the thermal wave cycle, wet cooling
towers, high evaporation temperatures and evacuated tube collectors produced the
highest COP values. Moreover, the heat capacity of the adsorbent bed and its shell
should be low for the simple and heat recovery cycles and the adsorbent bed should
be cooled down to the condensation temperature for all cases to achieve the highest
possible COP. The selection of working pair should depend on the temperature of the
available heat source (solar energy in this study) since each working pair has a dis-
v
tinct operating temperature range. Furthermore, there is always a need for backup
power for the analyzed location and the system.

Keywords: Adsorption, Cooling, Modeling, Simulation, Solar
vi
Z
GNE ENERJS DESTEKL ADSORPSYONLU DEAL SOUTMA
SSTEMLERNN SAYISAL MODELLENMES VE BAARIM
NCELEMES

Taylan, Onur
Yksek Lisans, Makina Mhendislii Blm
Tez Yneticisi : Do. Dr. Derek K. Baker
Ortak Tez Yneticisi: Prof. Dr. Bilgin Kaftanolu

Mays 2010, 139 sayfa

Btn dnyada enerji tketimi gn getike artmakta ve bu durum aratrmaclar
srdrlebilir zmler bulmaya sevk etmektedir. Soutmaya olan talep de artan
enerji talebinin nedenlerinden biridir. Bu aratrma soutma iin artan enerji talebini
azaltmann srdrlebilir yollarndan biri olan gne enerjisi destekli soutma
sistemleri zerinde younlamtr. Bu almada gne enerjisi destekli soutma
sistemlerinin genel teorik baarm eilimleri TRNSYS ve MATLAB yazlmlar
kullanlarak incelenmitir. eitli dng iyiletirmeleri, adsorban-soutucu ifti,
ortam ve tasarm koullarnn performans zerindeki etkileri bir dizi zamandan
bamsz veya zamana bal benzetim yoluyla aratrlmtr. Bu etkenlere ek olarak
normalletirilmi bir model sunularak sistem boyutu, ek g ihtiyac, gne topla
alan ve adsorban ktlesinin etkileri de incelenmitir. Sonular baarm katsaylar ve
onlarn oranlar eklinde verilmitir. ncelenen durumlara gre termik dalga dngs,
ya soutma kulesi ve yksek buharlama scakl yksek baarm katsay ortaya
karmtr. Ayrca basit ve sl kazanml dng iin s sas dk adsorban yatak
ve kabuu kullanmak ile her dng iin adsorban yata younlama scaklna
drmek baarm katsaysn artrmtr. Adsorban-soutucu ifti seimi ise her
iftin alma scaklnn farkl olmasndan dolay mevcut s kaynann (bu
vii
alma iin gne enerjisinin) ulaabilecei scakla baldr. Son olarak incelenen
blge ve her durum iin daima bir ek g ihtiyac olduu sonucuna varlmtr.

Anahtar Kelimeler: Adsorpsiyon, Soutma, Modelleme, Benzetim, Gne Enerjisi
viii
To Mom and Dad
ix
ACKNOWLEDGEMENTS
Firstly, I would like to state my heartfelt appreciation to my supervisor Assoc. Prof.
Dr. Derek K. Baker for his perceptive supervision, everlasting support and encour-
agement. I appreciate his continuous help inside and outside the academia within the
last three years.
I also would like to thank my co-supervisor Prof. Dr. Bilgin Kaftanolu. I feel very
privileged to have his precious advices, guidance and leadership.
Additionally, I would like to express my deepest feelings to my dearest parents for
their full support on my decisions at every step of my life. At this point, it is so much
momentous than ever to have them by my side. I am forever indebted to them.
I would like to show gratitude to my colleagues for their sincere friendships and
technical support throughout the thesis period. Thesis period has become so much
consequential with their harmony.
Finally, I would like to thank Assoc. Prof. Dr. Cemil Yamal, Mr. smail Solmu and
Mr. Ahmet alar for their insightful discussions and supports.
x
TABLE OF CONTENTS
ABSTRACT ................................................................................................................ iv
Z ............................................................................................................................... vi
ACKNOWLEDGEMENTS ........................................................................................ ix
TABLE OF CONTENTS ............................................................................................. x
LIST OF TABLES .................................................................................................... xiii
LIST OF FIGURES .................................................................................................. xiv
LIST OF SYMBOLS ............................................................................................... xvii
CHAPTER
1. INTRODUCTION ............................................................................................... 1
1.1 Background Information.............................................................................. 1
1.2 Current Solar-Thermal Power Situation ...................................................... 3
1.3 Overview of Adsorption Cooling Systems .................................................. 4
1.4 Motivation of the Present Study .................................................................. 5
2. SURVEY OF LITERATURE AND OBJECTIVES ........................................... 7
2.1 Overview of Cooling Cycles ....................................................................... 7
2.1.1 Vapor-Compression Cooling Cycle ..................................................... 7
2.1.2 Absorption Cooling Cycles .................................................................. 9
2.1.3 Desiccant Cooling .............................................................................. 11
2.2 Adsorption Cooling Cycles ....................................................................... 12
2.2.1 Definitions .......................................................................................... 12
2.2.2 Previous Experimental Studies ........................................................... 19
2.2.3 Previously Developed Numerical Modeling ...................................... 23
xi
2.2.4 Previous Theoretical Studies .............................................................. 35
2.3 Adsorption Capacity Models ..................................................................... 40
2.3.1 Previous Studies on Working Pair Comparison ................................. 43
2.4 TRNSYS Software .................................................................................... 45
2.5 Objectives of the Present Study ................................................................. 46
3. MODELS ........................................................................................................... 48
3.1 Modifications on Thermal Wave Adsorption Cooling Cycle Model ........ 48
3.1.1 Thermal Wave Adsorption Cooling Cycle with Adiabatic Mass
Recovery (AMR) ................................................................................ 49
3.1.2 Thermal Wave Adsorption Cooling Cycle with Isothermal Mass
Recovery (IMR) .................................................................................. 52
3.2 Solar Thermal System Model .................................................................... 53
3.2.1 Solar Thermal Collectors .................................................................... 54
3.2.2 Heat Exchanger .................................................................................. 57
3.2.3 Circulation Pump ................................................................................ 58
3.2.4 Control Unit ........................................................................................ 59
3.2.5 Weather Data ...................................................................................... 59
3.2.6 Integration of Solar Thermal System with Adsorption Cycle
Models ................................................................................................ 59
3.3 Normalized Model ..................................................................................... 61
4. ANALYSES AND VERIFICATION ............................................................... 64
4.1 Analyses..................................................................................................... 64
4.2 Error Quantification ................................................................................... 69
4.3 Solar Thermal Collector Model Verification ............................................. 71
5. RESULTS .......................................................................................................... 74
5.1 Adsorption Cooling Cycle Comparison .................................................... 74
xii
5.2 Adsorbent Refrigerant Pair Comparison ................................................ 77
5.3 Design and Working Conditions Comparison ........................................... 81
5.3.1 Effects of Collector Choice and Solar Radiation Level ..................... 81
5.3.2 Effects of Condensation Temperature ................................................ 83
5.3.3 Effects of Evaporation Temperature .................................................. 86
5.3.4 Effects of Excess Bed Temperature ................................................... 89
5.3.5 Effects of Heat Capacity Ratio ........................................................... 93
5.3.6 Effects of Maximum Bed Temperature .............................................. 95
5.3.7 Summary of Effects of Investigated Parameters ................................ 96
5.4 Normalized Model Results ........................................................................ 98
6. DISCUSSIONS ............................................................................................... 105
7. CONCLUSIONS ............................................................................................. 110
8. SUGGESTIONS FOR FUTURE INVESTIGATIONS .................................. 112
BIBLIOGRAPHY .................................................................................................... 115
APPENDICES
A. SAMPLE TRNSYS INPUT (DECK) FILE ................................................... 125
B. SAMPLE TRNSYS STUDIO REPORT ........................................................ 135

xiii
LIST OF TABLES
TABLES
Table 2.1. Adsorbent and adsorbed refrigerant masses for the back and front
regions of the bed. ............................................................................................... 27
Table 2.2. COP comparison of adsorption systems with different configurations ..... 37
Table 2.3. Coefficients for D-A equations. ................................................................. 41
Table 2.4. Coefficients for adsorption capacity model of zeolite NaX water
pair. ................................................................................................................... 42
Table 2.5. Constants for adsorption cycle models. ..................................................... 43
Table 3.1. Performance coefficients for the flat plate solar collectors. ...................... 55
Table 3.2. Other necessary parameters for the flat plate solar collectors. .................. 55
Table 3.3. Incidence angle modifiers for the flat plate collectors. ............................. 56
Table 3.4. Parameters for the evacuated tube solar collector. .................................... 57
Table 3.5. Parameters for the heat exchanger. ............................................................ 58
Table 3.6. Parameters for the circulation pump. ......................................................... 58
Table 4.1. Investigated conditions. ............................................................................. 66
Table 4.2. Analyzed maximum bed temperature ranges for each pair. ...................... 67
Table 4.3. Simulation durations and 1
st
Law error comparison of constant and
variable specific heats. ........................................................................................ 70
Table 4.4. Solar collector verification results for FPL. .............................................. 71
Table 4.5. Solar collector verification results for FP. ................................................. 72
Table 4.6. Solar collector verification results for ET. ................................................ 72
xiv
LIST OF FIGURES
FIGURES
Figure 2.1. Vapor-compression cooling system. ......................................................... 8
Figure 2.2. Simple absorption cooling cycle. ............................................................ 10
Figure 2.3. Basic desiccant cooling cycle. ................................................................. 11
Figure 2.4. (a) Schematic and (b) Clapeyron diagram for an ideal simple
adsorption cycle. ................................................................................................. 13
Figure 2.5. Clapeyron diagram for mass recovery cycle. .......................................... 17
Figure 2.6. Schematic of the heat transfer fluid (HTF) flow for a half cycle in a
thermal wave cycle. ............................................................................................. 18
Figure 2.7. Schematic presentation of the convective thermal wave cycle. .............. 19
Figure 3.1. Clapeyron diagram for back region of thermal wave adsorption
cycles. ................................................................................................................. 50
Figure 3.2. Clapeyron diagram for front region of thermal wave adsorption
cycles. ................................................................................................................. 50
Figure 3.3. Solar thermal system and its integration with the adsorption cycle. ....... 53
Figure 3.4. TRNSYS screen capture of the solar thermal system. ............................ 54
Figure 3.5. Integration of solar thermal system with adsorption cycle in
TRNSYS. ............................................................................................................ 60
Figure 5.1. Comparison of different adsorption cycle types for Z1 pair, T
cond
=
30
o
C, T
evap
= 10
o
C, R = 10 and AT
excess
= 0
o
C. .................................................... 75
Figure 5.2. Comparison of different adsorption cycle types for CA pair, T
cond
=
30
o
C, T
evap
= 10
o
C, R = 10 and AT
excess
= 0
o
C. .................................................... 76
Figure 5.3. Clapeyron diagrams for different working pairs with T
hot
= 90
o
C,
T
cond
= 30
o
C, T
evap
= 10
o
C, R = 10 and AT
excess
= 0
o
C. ......................................... 78
hot
= 90
o
C,
T
cond
= 35
o
C, T
evap
= 5
o
C, R = 10 and AT
excess
= 0
o
C. ........................................... 78
xv
Figure 5.5. Working pair comparison using different adsorption cycle types
with T
cond
= 30
o
C, T
evap
= 10
o
C, R = 10 and AT
excess
= 0
o
C ................................. 79
Figure 5.6. Working pair comparison using different adsorption cycle types
with dry cooling tower, T
evap
= 10
o
C, R = 10 and AT
excess
= 0
o
C. ....................... 80
Figure 5.7. Collector and solar radiation level comparison for simple cycle, Z1
pair, T
cond
= 20
o
C, T
amb
= 35
o
C, T
evap
= 10
o
C, R = 10 and AT
excess
= 10
o
C. .......... 82
Figure 5.8. Collector and solar radiation level comparison for simple cycle, Z1
pair, T
cond
= 20
o
C, T
amb
= 35
o
C, T
evap
= 10
o
C, R = 10 and AT
excess
= 10
o
C. .......... 83
Figure 5.9. Condensation temperature comparison for different working pairs,
NMR, T
evap
= 10
o
C, R = 10 and AT
excess
= 10
o
C .................................................. 84
Figure 5.10. Cooling tower comparison for CA and Z1 pairs, simple and heat
recovery cycles, T
evap
= 10
o
C, R = 10 and AT
excess
= 10
o
C .................................. 85
Figure 5.11. Cooling tower comparison for different adsorption cycle types,
working pairs, T
evap
= 10
o
C, R = 10 and AT
excess
= 0
o
C ....................................... 86
Figure 5.12. Evaporation temperature comparison for NMR, ZW, CA and SG
pairs, T
cond
= 30
o
C, R = 10 and AT
excess
= 10
o
C ................................................... 87
Figure 5.13. Evaporation temperature comparison for simple and heat recovery
cycles, CA and Z1 pairs, T
cond
= 30
o
C, R = 10 and AT
excess
= 10
o
C ..................... 88
Figure 5.14. Evaporation temperature comparison for different adsorption
cycle types and working pairs, T
cond
= 30
o
C, R = 10 and AT
excess
= 0
o
C ............. 89
Figure 5.15. Excess bed temperature comparison for NMR and ZW, CA and
SG pairs, T
cond
= 30
o
C, T
evap
= 10
o
C and R = 10 ................................................. 90
Figure 5.16. Excess bed temperature comparison for simple and heat recovery
cond
= 30
o
C, T
evap
= 10
o
C and R = 10 ........................ 91
Figure 5.17. Excess bed temperature comparison for different adsorption cycle
types and working pairs, T
cond
= 30
o
C, T
evap
= 10
o
C and R = 10 ......................... 92
Figure 5.18. Heat capacity ratio comparison for NMR, ZW, CA and SG pairs,
T
cond
= 30
o
C, T
evap
= 10
o
C and AT
excess
= 0
o
C ...................................................... 93
Figure 5.19. Heat capacity ratio comparison for simple and heat recovery
cond
= 30
o
C, T
evap
= 10
o
C and AT
excess
= 10
o
C. ........... 94
xvi
Figure 5.20. Heat capacity ratio comparison for different adsorption cycle
types and working pairs, T
cond
= 30
o
C, T
evap
= 10
o
C and AT
excess
= 0
o
C. ............. 96
Figure 5.21. Effects of investigated parameters on COP
ads
for IMR and Z1 pair ..... 97
sys,clg
for different
adsorption cycle types and ZW pair. ................................................................... 99
Figure 5.23. f vs. q
Storage,max
for various collector types and T
hot
, S = 1, simple
cycle, wet cooling tower, R = 10, AT
excess
= 10
o
C ............................................. 100
Figure 5.24. l vs. q
Storage,max
hot
, S = 1, simple
cycle, wet cooling tower, R = 10, AT
excess
= 10
o
C ............................................. 101
Storage,max
for various S values, flat plate collector, simple
cycle, T
hot
= 90
o
C, wet cooling tower, and R = 10. ........................................... 101
Storage,max
for various S values, flat plate collector, simple
cycle, T
hot
= 90
o
C, wet cooling tower, and R = 10. ........................................... 102
Figure 5.27. A
coll,N
vs. T
hot
(
o
C) for different collector, adsorption cycle, cooling
tower types and R and AT
excess
= 0
o
C. ................................................................ 103
Figure 5.28. m
ads,N
vs. T
hot
(
o
C) for evacuated tube collector, simple and heat
recovery cycles, different cooling towers and T
o
and R = 10. .......................... 104
xvii
LIST OF SYMBOLS
a, b, b
o
coefficients in adsorption capacity relation of zeolite NaX water pair
c specific heat of incompressible substance (kJ kg
-1
K
-1
)
c
p
ideal gas specific heat at constant pressure (kJ kg
-1
K
-1
)
d a coefficient in adsorption capacity relation of zeolite NaX water pair
d
o
, d
1
, d
2
performance coefficients in collector thermal efficiency equation
D a coefficient in Dubinin Astakhov equation
E a coefficient in adsorption capacity relation of zeolite NaX water pair
f solar fraction (-)
G solar radiation (W m
-2
)
h specific enthalpy (kJ kg
-1
)
H enthalpy (kJ)
h
fg
heat of vaporization / evaporation (kJ kg
-1
)
Ah
ads
heat of adsorption (kJ kg
-1
)
k a coefficient in Dubinin Astakhov equation
l loss fraction (-)
m mass (kg)
m mass flow rate (kg h
-1
)
n exponent of Dubinin Astakhov equation
ns coefficients in adsorption capacity relation of zeolite NaX water pair
P pressure (kPa)
q heat transfer per unit mass of adsorbent (kJ kg
-1
)
Q heat transfer (kJ)
r coefficient of performance ratio (-)
R ratio of beds design to inherent heat capacities (-)
S user defined size of the cooling system (-)
t time (min)
T temperature (
o
C)
T
hot
maximum bed temperature (
o
C)
AT
excess
excess bed temperature (T
o
T
cond
) (
o
C)
u specific internal energy (kJ kg
-1
)
U internal energy (kJ)
X adsorption capacity (kg
ref
/ kg
ads
)
AX adsorption capacity change for a half cycle (kg
ref
/ kg
ads
)
y thermal wave position (-)
Greek Symbol

c convergence criterion of numerical iteration scheme
q thermal efficiency

xviii
Subscripts

a common state after adiabatic mass recovery
ads adsorbent / adsorption cycle
amb ambient
b back region of the bed
base base case
bed adsorbent bed
clg cooling capacity
clr cooler
coll solar thermal collector
cond condensation
evap evaporation
f front region of the bed
F fuel
flow flowing, not stagnant heat transfer fluid
hex heat exchanger in solar thermal system
htf heat transfer fluid
htr heater
i common state after isothermal mass recovery
in inlet
liq liquid refrigerant
load cooling load
m low-pressure adsorbent bed
max maximum
min minimum
n high-pressure adsorbent bed
N normalized
o minimum of bed
out exit / outlet
P product
rec heat recovery / recovered
ref adsorbed refrigerant
rev reversible
rfrc reference
s unspecified / unknown
sat saturation
shell adsorbent bed shell
sys solar thermal system integrated adsorption cooling system
tot total
v refrigerant vapor
w thermal wave
wb wet bulb
Superscript

+ immediately after corresponding symbol
xix
Abbreviations

AMR adiabatic mass recovery
CA activated carbon ammonia pair
COP coefficient of performance
CM activated carbon methanol pair
D-A Dubinin Astakhov equation
ET evacuated tube solar thermal collector
FP, FPL flat plate solar thermal collector
HRec heat recovery cycle
HTF heat transfer fluid
IAM incidence angle modifiers
IMR isothermal mass recovery
SG silica gel water pair
NMR no mass recovery
TW thermal wave cycle
WP working pair
Z1 zeolite NaX water pair
ZW zeolite X13 water pair

1
CHAPTER 1

INTRODUCTION
1.1 Background Information
World marketed energy consumption is projected to increase by 44% from 2006 to
2030 according to the U.S. Energy Information Administration [1]. With the in-
creasing trend in energy consumption and worldwide economic growth, the general
trend in cooling and air conditioning requirements of industry and buildings is also
increasing. To meet the demand in cooling, mechanical vapor-compression systems
are commonly used which can be classified as conventional systems [2]. These sys-
tems are very popular due to their high coefficients of performance, small sizes and
low weights. However, they also exhibit some disadvantages such as contributing to
global warming and ozone layer depletion and high energy consumptions [3].
One of the disadvantages of conventional vapor-compression systems can be stated
that these systems include refrigerants such as chlorofluorocarbon (CFC), hydro-
chlorofluorocarbon (HCFC) or hydrofluorocarbons (HFC) which have high global
warming potential and ozone depletion potential [4-6]. Since the global warming
problem is presently more critical, researchers are studying ways to reduce the emis-
sion of these greenhouse gases and overcome this significant disadvantage of con-
ventional cooling systems, especially after the Montreal (1988) and Kyoto (1998)
conventions. These conventions are held to force the participant countries to limit the
greenhouse gas emissions. In Montreal protocol (1988), some severe regulations
were agreed to reduce ozone layer depletion. In Kyoto protocol (1998), new regula-
tions on the CFC, HCFC and HFC emissions have been agreed to reduce greenhouse
gas emissions.
2
Another main disadvantage of conventional cooling systems is their large electricity
consumption which causes a need for new investments and new infrastructure, such
as new power plants, transmission and distribution lines on the electricity networks
[6]. Although building new power plants seems to be a valid solution, each power
plant requires fuels, and this requirement causes depletion in resources. In general, a
rapid growth in energy consumption comes with a rapid depletion in resources.
Therefore, the countries which do not have enough national resources start to import
the necessitated resources or buy electricity directly from their neighbors which cre-
ates issues about national security and add extra load on economies, especially for
countries like Turkey.
Turkey is a developing country and like other developing countries has a rapidly
growing energy demand. According to statistical data of the Turkish Electricity
Transmission Company (TEA), the average demand in electricity increased annu-
ally by 8.3% for the period of 1975-2008 [7]. In 2007, about 81% of the electricity
demand of Turkey was met by thermal sources (coal, lignite, fuel oil, LPG and natu-
ral gas), of which 61.2% was natural gas which was mainly imported from nearby
countries [8]. The remaining 19% of the electricity supply consisted of hydro dams,
geothermal, wind, etc. According to TEA and under one scenario, electricity de-
mand in Turkey will exceed the supply in 2016-2017 [9].
One of the main constituents of the increasing energy demand in Turkey is increasing
cooling loads, especially on the Mediterranean coast of Turkey due to long and hot
summers [10]. For instance, Antalya is situated on the Mediterranean coast of Turkey
and has electricity shortages and blackouts in summer due to large air conditioning
loads and consequent large loads on electricity networks. This city is also character-
ized by a large tourism industry. There are several top rated resorts, hotels, etc., and
these centers use mainly conventional vapor-compression air conditioners to meet
the cooling demand which also contributes to the significant load on the electricity
networks. According to governor of Antalya, in 2007, daily peak electricity con-
sumption of Antalya was around 18 million kWh, while the production in Antalya is
3
around 3 million kWh [11]. For all these reasons, Antalya needs to produce electrici-
ty or reduce its electricity demand.
As a result, both worldwide and nationwide, there is a strong need for alternative
cooling technologies that use sustainable and renewable energy supplies. The proper
technology should be environmentally benign and provide high performance so that
it can be comparable with the conventional and commonly-used vapor-compression
cooling systems.
1.2 Current Solar-Thermal Power Situation
According to United Nations worldwide estimates, practical hydroelectric resources
are less than 0.5 TW while tides and ocean movements can create cumulative power
less than 2 TW. The geothermal energy of all continents is approximately 12 TW of
which only a small amount of can be extracted. Additionally, potential wind power
and other non-solar renewable energy sources are estimated as equivalent 2-4 TW.
Among all discovered renewable energy sources, solar energy by far has the highest
potential with an average of 120,000 TW that hits the earths ground [12].
As all other renewable energy sources, usage of solar thermal energy is increasing. In
2006, there was an increase of 22% in new installations throughout the world when
compared to the previous year according to the International Energy Agencys report
of 2008 [13]. This increase corresponds to 18.3 GW of new capacity and 26.1 million
square meters [13]. Especially, when flat plate and evacuated tube collectors are con-
sidered, China and Taiwan have the highest installed thermal collector area which
corresponds to approximately 80 GW of thermal power whereas Turkey had approx-
imately 7.1 MW of installed thermal collector power in 2007 [14].
As mentioned by Abu Hamdeh and Al-Muhtaseb, the Mediterranean countries may
save 40-50% of their energy used for air-conditioning by implementing solar-driven
air-conditioning systems [15]. Like other Mediterranean countries, Turkey has also a
large opportunity for solar energy. The Mediterranean coast of Turkey is a popular
4
place for summer vacations due to its long, hot and sunny summer season. Antalya is
situated on the Mediterranean coast of Turkey and is well-known for its luxury ho-
tels. These hotels often use conventional vapor compression air-conditioning systems
that require electricity and contribute to Antalyas annual peak electrical demand
occurring during the summer [16]. Therefore, it is essential to reduce the electricity
load on the network, especially summer cooling loads.
1.3 Overview of Adsorption Cooling Systems
One way to decrease electricity demand due to cooling is to use environmentally
benign, thermally-powered cooling systems such as adsorption systems where the
term adsorption in this study refers to solid physical sorption, not chemical sorp-
tion. Basically, in an adsorption cooling cycle (or heat pump), the mechanical com-
pressor in the well-known conventional vapor-compression air conditioners that is
generally powered by electricity is replaced with a thermal compressor that is driven
by low grade thermal energy like solar energy or waste heat, and they do not require
electricity, except for circulation pumps. Although adsorption cooling systems are
not widely available in the market, these systems are proposed as a promising tech-
nology that deserves further research since they do not include rotating compressor
parts [17-19].
Some of the advantages of adsorption cooling systems relative to conventional va-
por-compression systems can be listed as [17-19]:
- Environmentally benign
- Can be operated with low-grade thermal energy such as solar, geothermal and
waste heat
- Thermal energy storage possible
- Does not have moving parts, therefore long lifetime without noise and vibra-
tion
- Simple control and maintenance
5
Nevertheless, there are also some disadvantages of adsorption cooling systems com-
pared with the conventional vapor-compression systems that can be listed as [17-19]:
- Low coefficient of performance
- Intermittent (not continuous) cooling
- System has to be operated under vacuum conditions and it is hard to maintain
this vacuum
- Larger volume and weight
1.4 Motivation of the Present Study
The general goal is to find a solution to meet the rapidly growing demand for energy
through an alternative cooling technology, specifically, adsorption cooling systems.
This study, in particular, looks at the general performance trends of the adsorption
cooling systems coupled with solar thermal systems, which are referred to as solar-
thermal powered adsorption cooling systems, as operating conditions and several
design parameters are varied. By common use of the adsorption cooling systems, it is
believed that it would be possible to reduce the electricity demand due to cooling in
areas like the Mediterranean coast of Turkey. Like other Mediterranean cities of Tur-
key, Antalya has large opportunities for solar energy. Systems like solar-powered
adsorption cooling cycles can be used to reduce the load on electrical power by mak-
ing the air conditioning independent of electricity. This study also investigates the
suitability of the adsorption systems for the cities on the Mediterranean coast of Tur-
key, specifically Antalya. This study is done to assess the feasibility using solar en-
ergy for thermally-driven adsorption cooling systems.
Additionally, this study not only provides general guidelines to improve the perfor-
mances of the adsorption cooling systems, it also reduces the time and cost of exper-
iments by providing broad directions to pursue. This study foresees the difficulties
and complexities of the experimental work by comparing and contrasting several
cases by means of simulations.
6
Within this study, background information on the current energy situation, solar ap-
plications and adsorption systems are introduced in Chapter 1. In Chapter 2, an over-
view of refrigeration cycles and the completed studies on adsorption cooling systems
are given. Newly introduced models are presented in Chapter 3 with the correspond-
ing mathematical derivations. In Chapter 4, analyzed cases and model verifications
are given. Detailed results of the analyzed cases are presented in Chapter 5 while the
discussions on these results are given in Chapter 6. Finally, conclusions are presented
in Chapter 7, and some suggestions as future work are provided in Chapter 8.
7
CHAPTER 2

SURVEY OF LITERATURE AND OBJECTIVES
2.1 Overview of Cooling Cycles
The adsorption cooling cycle is just one of many cooling cycles. To put this study
into a broader context, a brief overview of the main refrigerant cycles except for the
adsorption cycle is provided, while a more detailed description of the adsorption cy-
cle is given in Section 2.2. The cooling cycles presented in this section are vapor-
compression, absorption cycles and desiccant cooling cycles.
2.1.1 Vapor-Compression Cooling Cycle
As stated before and according to Moran and Shapiro, vapor-compression cooling
cycles are the most commonly used cooling cycles [20]. A vapor-compression cycle
works between two thermal reservoirs: one reservoir can be considered as the cold
reservoir (or the space to be cooled) and the other as the hot reservoir (or the sur-
roundings). A refrigerant is circulated within the system whose components are giv-
en in a schematic in Figure 2.1. As shown in this figure, this cooling cycle consists of
a mechanically-driven compressor, a condenser, an evaporator and an expansion
valve. The compressor in the vapor-compression cooling cycle is mechanically driv-
en so that it should be operated externally by means of an electric motor or similar.
In general, the ideal vapor-compression cooling cycles follow the following four
processes as shown in Figure 2.1 [20, 21]:
- Process 12: The refrigerant at a saturated vapor phase enters the compres-
sor and undergoes an adiabatic compression process where work input is nec-
essary.
8
- Process 23: The refrigerant at a superheated vapor state undergoes an iso-
baric condensation process where heat is transferred from the condenser to
the high temperature reservoir.
- Process 34: The saturated liquid refrigerant is throttled in an isenthalpic
expansion process.
- Process 41: The refrigerant at a two-phase liquid-vapor mixture state un-
dergoes an isobaric evaporation process where heat transfer from the cold
reservoir occurs to the evaporator.

Figure 2.1. Vapor-compression cooling system [20].

One of the main advantages of the vapor-compression cycle is that this cycle can be
used for freezing purposes since the refrigerants can be operated at temperatures be-
low their freezing points. Another main advantage of this cycle is its high coefficient
of performance as stated in Section 1.1. The actual vapor-compression cooling cycle
9
may deviate from the actual cycle by having the refrigerant at state 1 as a superheat-
ed vapor, heat transfer from the refrigerant in the compressor, pressure drop in the
condenser, sub-cooled refrigerant at state 3, more heat transfer to the refrigerant in
the evaporator, etc. For more information, interested readers are referred to Moran
and Shapiro [20] and Sonntag et al. [21].
2.1.2 Absorption Cooling Cycles
Absorption cycles are also referred as liquid sorption cycles since absorbent is in the
liquid phase in the absorbent bed (or absorber). Additionally, like adsorption cooling
cycles, absorption cooling cycles are heat-powered cooling cycles and do not require
mechanical compressors that have moving components. Unlike conventional vapor-
compression cycles which operate between two thermal reservoirs, absorption cool-
ing cycles operate between three thermal reservoirs: one reservoir is the cold reser-
voir (or the space to be cooled), the second reservoir is at the temperature of the
available heat source and the last reservoir is the surroundings (or the reservoir to
which heat is rejected from the condenser). Another difference between vapor-
compression and absorption cooling cycles is that in vapor-compression cycles, re-
frigerant vapor is compressed in the mechanically-driven compressor, which reduces
the specific volume of the refrigerant significantly and therefore requires a large
amount of work. Conversely, in absorption cooling cycles, refrigerant in the liquid
phase undergoes a compression process and the required amount of work is consid-
erably reduced [22]. However, the overall energy used to generate refrigerant in the
absorption cooling cycle is higher than the energy used in the vapor-compression
cycle, so that the coefficient of performance of absorption cooling cycle is smaller
[23]. Additionally, absorption systems are not suitable for mobile applications since
they involve liquid absorbents, and these systems are susceptible to corrosion if they
are operated under high temperatures (above 200
o
C for instance) [24]. According to
Meunier, another disadvantage of absorption cooling cycles is their high initial costs
and they also require wet cooling towers (or low temperatures for the condenser)
[25].
10
Two well-known absorbent refrigerant (working) pairs are lithium bromide (LiBr)
water (H
2
O) and ammonia (NH
3
) water (H
2
O) pairs. A schematic of a simple
(without enhancements) absorption cooling cycle is given in Figure 2.2, and the de-
tails of this cycle can be found in the books of Trott and Welch [23], Alefeld and
Radermacher [26] and Herold et al. [27]. Basically, refrigerant at low pressure is
absorbed and forms a strong liquor (or solution) before it is directed to the generator.
After the liquor leaves the absorber, pressure is increased by a pump and refrigerant
vapor is generated from liquor in the generator through heat transfer. Subsequently,
the refrigerant vapor undergoes the common cooling cycle (condenser, throttling
valve and evaporator) while the weak solution is directed to the absorber again.

Figure 2.2. Simple absorption cooling cycle [23].
Moreover, absorbent pairs generally have lower temperature swing than adsorbent
pairs, and absorption cooling necessitates use of liquid-gas heat exchangers rather
than solid-gas heat exchangers found in adsorption cooling [25].
11
2.1.3 Desiccant Cooling
Another alternate cooling cycle that can be run with a low-grade thermal energy
source is desiccant cooling. In desiccant cooling cycles, desiccants are used to dehu-
midify the air. The weight of the wetted commercial desiccants can reach up to 11
times their dry weights [28]. By heating the saturated desiccant, moisture can be re-
moved thereby regenerating the desiccant.
The main advantages of desiccant cooling cycles over conventional vapor-
compression cycles are the use of a low-grade energy source so that exergy con-
sumption is lowered, no greenhouse gases are emitted and indoor air quality is im-
proved through ventilation. Additionally, desiccant cooling cycles can operate with
air as refrigerant and water for evaporative cooling.
The desiccant cooling cycles can be categorized into solid and liquid, or open and
closed desiccant cooling cycles. In open desiccant cooling cycles, the ambient air
enters the cycle and is then dehumidified by a desiccant (e.g., a bed packed with des-
iccants) as shown in Figure 2.3.

Figure 2.3. Basic desiccant cooling cycle [29].

12
The moisture (water vapor) is adsorbed (for solid desiccant cooling cycles) or ab-
sorbed (for liquid desiccant cooling cycles) in the desiccant. There occurs a tempera-
ture rise due to adsorption/absorption, and then sensible cooling occurs in a heat ex-
changer (heat recovery as labeled in Figure 2.3). Before entering the room to be
cooled (or ventilated), the air is subjected to evaporative cooling through humidifiers.
Simultaneously, the warm air to be exhausted is cooled evaporatively, and then it
passes through a sensible heat exchanger as it recovers some amount of the heat of
the inlet stream. Finally, the exhaust stream is heated by a low-grade thermal energy
before regenerating the desiccant in the wheel. Alternatively, the air can be circulated
between the inlet stream and the exhaust stream to form a closed desiccant cooling
cycle. More details can be found in Waugaman et al. [28], Henning et al. [29], Daou
et al. [30] and ASHRAEs publication [31].
2.2 Adsorption Cooling Cycles
Although the main attention in this study is given to the cooling cycles, the defini-
tions in this section are also valid for other adsorption cycles (heat pumps, ice-
making, etc.) as well.
2.2.1 Definitions
In the literature, the adsorption cooling cycle is well-established and several review
papers have been published (e.g., [17, 19, 32-34]) describing the cycles in this sec-
tion.
a. Simple adsorption cooling cycle
As mentioned previously, in an adsorption cooling cycle, the mechanical compressor
in the well-known vapor compression cooling cycle that is generally powered by
electricity is replaced with a thermal compressor that can be driven using low grade
thermal energy like solar energy or waste heat. Simple (adsorption) cycle herein re-
fers to an adsorption cycle with a spatially, but not temporally isothermal and isobar-
13
ic adsorbent bed. In the literature, the term simple is generally omitted or instead
the term intermittent is used. A schematic of an ideal simple adsorption cycle is
shown in Figure 2.4. The cycle is characterized by temporal variations in the refrig-
erants vapor pressure and the adsorbents temperature and adsorption capacity (X),
which is defined as the ratio of adsorbed refrigerant mass to adsorbent mass. The
ideal simple cycle is composed of four processes as shown in Figure 2.4.
- Process 12: Isosteric (constant X) heating
- Process 23: Isobaric heating and desorption
- Process 34: Isosteric (constant X) cooling
- Process 41: Isobaric cooling and adsorption

Figure 2.4. (a) Schematic and (b) Clapeyron diagram for an ideal simple adsorption
cycle [35].
At the beginning of the adsorption cycle, the maximum amount of refrigerant is ad-
sorbed on the adsorbent (X
max
) in the adsorbent bed which is at the minimum bed
temperature (T
o
) and evaporator pressure (P
evap
) (state 1). Additionally, all valves are
14
closed. In process 12, the adsorbent bed is heated by the driving heat transfer (q
F
)
from a hot thermal reservoir. The vapor pressure increases while the adsorption ca-
pacity essentially remains constant at the maximum adsorption capacity (isosteric
process, dX = 0). Process 12 ends when the adsorbent bed reaches the condensa-
tion pressure (P
cond
) (state 2).
At the state 2, the valve connecting the adsorbent bed to the condenser that is shown
in Figure 2.4 is opened allowing desorbed refrigerant to flow to the condenser. For
process 23 the bed continues to be heated using q
F
and desorption and condensa-
tion occur isobarically at P
cond
. Heat transfer occurs from the condenser to the envi-
ronment. Process 23 ends when the bed reaches its maximum temperature (T
hot
)
and minimum adsorption capacity (X
min
) (state 3).
Afterwards, the valve between the bed and the condenser is closed and the bed is
brought into thermal communication with the environment allowing heat transfer
from the adsorbent bed to the environment (q
bed
) to occur. The bed cools isosterically
for process 34 until the pressure inside the bed decreases to P
evap
(state 4).
The valve connecting the evaporator to the bed is opened and the bed continues to be
cooled to T
o
at constant P
evap
causing refrigerant to be adsorbed. This adsorption pro-
cess pulls saturated liquid refrigerant from the condenser through the throttling de-
vice and evaporator creating the cooling by means of the product heat transfer (q
P
) to
the evaporator. The process ends (and the cycle is complete) when the adsorbent bed
reaches T
o
(state 1).
b. Adsorption cooling cycle with heat recovery
Heat recovery cycle herein refers to an adsorption cooling cycle with two spatially
isothermal beds and ideal heat recovery. Each bed follows the same four processes
(12341) given in Figure 2.4b, and these two beds are operated out of phase
by 180
o
. Therefore, while one bed is being heated from T
o
to T
hot
, the other adsorbent
bed is being cooled from T
hot
to T
o
. When these two beds are thermally connected to
15
each other, they can reach thermal equilibrium at a common temperature (T
rec
)
through q
rec
that is heat transfer from the bed that is being cooled (process 31 in
Figure 2.4b) to the bed that is being heated (process 13 in Figure 2.4b). Therefore,
the total required q
F
is decreased by q
rec
and COP
ads
of the cycle is increased com-
pared to the simple cycle that is operating under same conditions (T
hot
, T
cond
, T
evap
,
etc).
In the design of heat recovery cycles, differences in sensible and latent heat capaci-
ties between the two analyzed adsorbent beds limit the possible thermal regeneration.
The effect of these differences become more apparent when the heat recovery pro-
cess is divided into four distinct sorption regimes [36]. These sorption regimes are
defined based on the common bed temperature, T
rec
.
- Only Adsorption Regime occurs when T
1
T
rec
< T
2
. The adsorbent bed that
is being heated is in the isosteric heating process (process 12 with dX = 0)
while the adsorbent bed that is being cooled is in the isobaric cooling and ad-
sorption process (process 41 with dX > 0).
- Only Desorption Regime occurs when T
4
T
rec
< T
3
is being heated is in the isobaric heating and desorption process (process 23
with dX < 0) while the adsorbent bed that is being cooled is in the isosteric
cooling process (process 34 with dX = 0).
- Paired Sorption Regime occurs when T
2
T
rec
T
4
is being heated is in the isobaric heating and desorption process (process 23
with dX < 0) while the adsorbent bed that is being cooled is in the isobaric
cooling and desorption process (process 41 with dX > 0).
- No Sorption Regime occurs when T
4
T
rec
T
2
. The adsorbent bed that is
being heated is in the isosteric heating process (process 12 with dX = 0)
while the adsorbent bed that is being cooled is in the isosteric cooling process
(process 34 with dX = 0).
Note that the paired sorption regime can occur only for the adsorbent bed pairs with
T
2
< T
4
, and no sorption regime can occur only for the adsorbent bed pairs with T
4
<
16
T
2
. Among the described four sorption regimes, the most suitable regime for heat
recovery is the paired sorption regime since the differences in heat capacities of the
adsorbent beds are similar in this regime. More details can be found in [37-39].
Heat recovery processes are also applied to multi-stage and cascading adsorption
cooling systems. These systems work under different temperature and/or pressure
levels which make it possible to recover heat. In multi-stage systems, the same work-
ing pair is used with several temperature and/or pressure levels and the heat of ad-
sorption or heat transfer from the condenser can be recovered in the same system.
Alternatively, in cascading systems, different working pairs are used to recover heat.
Cascading systems are especially advantageous over multi-stage systems if the oper-
ating temperature and/or pressure ranges of the working pairs are large [17].
c. Adsorption cooling cycle with mass recovery
Mass recovery can be classified as an enhancement process which is applied to the
simple adsorption cycles or before the heat recovery process in heat recovery adsorp-
tion cycles. The main purpose of the mass recovery is to ease the pressurization and
depressurization processes of the adsorbent beds. In two-bed adsorption cycles, one
adsorbent bed at state 1 and the other adsorbent bed at state 3 are physically connect-
ed to each other through a pipe to let the mass recovery process occur. The adsorbent
bed at state 3 is at high pressure and refrigerant vapor within this adsorbent bed
leaves and enters the other adsorbent bed at state 1 whose vapor pressure is lower.
This vapor refrigerant transfer between two adsorbent beds is referred to as a mass
recovery process, and ideally the mass recovery process ends when the two adsorbent
beds reach the same pressure level. As a result, a portion of the pressurization (pro-
cess 12) and depressurization (process 34) processes are completed without any
external source. Additionally, Clapeyron diagram changes slightly for cycles with
mass recovery as shown in Figure 2.5. Note that the adsorption swing also increases
with mass recovery. More information on mass recovery schemes can be found in
[37, 38, 40, 41].
17

Figure 2.5. Clapeyron diagram for mass recovery cycle (Legend: solid lines for mass
recovery cycle, dotted lines for simple cycle) (adapted from [37]).
d. Adsorption cooling cycle with thermal regeneration
Another commonly cited cycle enhancement scheme is thermal regeneration. Ad-
sorption cooling cycles with thermal regeneration can be divided into three main
types: thermal wave and convective thermal wave cycles and cycles with rotary ad-
sorbent beds.
In thermal wave adsorption cooling cycles, two adsorbent beds are thermally con-
nected to each other with the help of a heat transfer fluid (HTF). The heat transfer
fluid passes through each adsorbent bed and creates a temperature variation (wave)
inside the adsorbent beds. Due to the presence of this temperature wave inside the
bed, this cycle is referred to as a thermal wave (adsorption cooling) cycle. A temper-
ature schematic of the HTF loop for an ideal thermal wave cycle is shown in Figure
2.6 where the vertical position of the heat transfer fluid corresponds to its tempera-
ture. HTF that is heated to the maximum bed temperature (T
hot
) in a heater enters the
adsorbent bed that is to be heated and creates a temperature wave inside the adsor-
bent bed. Similarly, HTF that is cooled to the minimum bed temperature (T
o
) enters
the adsorbent bed which is being cooled and also creates a temperature wave inside
18
the adsorbent bed. The heater and cooler exit temperatures are fixed at T
hot
and T
o
.
As the cycle proceeds, sorption processes cause the beds exit temperatures to devi-
ate from T
hot
and T
o
as shown in Figure 2.6 (as oT
1
and oT
2
). These temperature
deviations result in the need for the heater and cooler. Each adsorbent bed basically
follows the same processes as the adsorbent beds in the simple adsorption cycle fol-
low, except the adsorbent beds in thermal wave cycle are not spatially isothermal.
The detailed information about each process in thermal wave adsorption cycle is giv-
en in Section 2.2.3.d with its mathematical model. Thermal wave adsorption cycle
was first designed and patented by Shelton [42, 43].

Figure 2.6. Schematic of the heat transfer fluid (HTF) flow for a half cycle in a ther-
mal wave cycle [44].
Another regenerative adsorption cooling cycle is the convective thermal wave cycle
proposed by Critoph [45] and patented by Sanburn [46]. In a convective thermal
wave cycle, instead of using an external HTF loop, refrigerant is used for the same
purpose. A schematic representation of the convective thermal wave cycle is given in
Figure 2.7. As can be observed from Figure 2.7, in addition to the adsorbent beds,
there are two inert beds that are packed with steel balls. These steel balls are used to
recover heat, storing the heat of adsorption when the adsorbent bed is in the adsorp-
tion process (processes 341) and this stored heat can be used when the adsorbent
bed is in desorption process (processes 123). The operation principle of the re-
maining part of the cycle is the same as in the thermal wave cycle.
19

Figure 2.7. Schematic presentation of the convective thermal wave cycle [45].
A different regeneration method is achieved by rotating the adsorbent beds. This cy-
cle is referred to as an adsorption (cooling) cycle with rotary beds. These rotary ad-
sorbent beds are coupled with condensers and evaporators, and this novel design was
suggested by Maier-Laxhuber and Kaubek [47] and further developed by Erickson
[48] and Ebbeson [49]. With this regenerative design, the aim is to provide continu-
ous cooling by having different temperature zones within the adsorbent bed and sim-
plify the process management in the regenerative adsorption cooling cycles. There-
fore, within a single adsorbent bed there are a number of sections at different temper-
ature levels. As the adsorbent bed rotates, each section of the bed undergoes the pro-
cesses one by one (from desorption to adsorption or vice versa). Within the rotary
adsorbent bed, air as a heat transfer fluid is passed through channels to supply or
remove the required heat transfers. The main concern of this design is the proper
adjustments of rotation speed and flow rates of the heat transfer fluid. More detailed
explanations can be found in [50-53].
2.2.2 Previous Experimental Studies
In the literature, there are many experimental studies on adsorption systems. Only the
studies that are related to the present study are chronologically summarized here for
20
brevity. Interested readers may refer to the provided references for more information
for each study.
One of the earliest studies which can be accessible on solar adsorption systems was
performed by Pons and Guilleminot in 1986 [54]. A solar powered adsorption ice
making system was analyzed from the fall 1984 to the end of summer 1985. Activat-
ed carbon methanol pair was selected, the maximum bed temperature was 95
o
C and
the evaporation temperature varied between -5
o
C and -10
o
C. A maximum COP of
0.15 for the system was obtained.
Another study was performed by Wang et al. in 1998 [55]. As stated in this study, the
most important parameters that affect the cycle performance are the cycle time and
maximum bed temperature for a simple adsorption cycle. This study investigated
these two parameters and compared experimental values with the simulated model
results. According to the experimental results, as the maximum bed temperature in-
creased, COP decreased and cooling power increased for activated carbon metha-
nol pair. Additionally, in this study, a new factor, packing coefficient, was defined to
incorporate the high difference between theoretical and experimental results.
Different sorption systems were compared thermodynamically by Pons et al. in 1999
in terms of COP using different working pairs [56]. The results stated that absorption
systems had higher COP values than adsorption systems and when cooling tower
types were compared, wet cooling towers increased the system performance com-
pared to dry cooling towers. Additionally, it was not possible to use water for ice
making (which required temperatures below 0
o
C) and results with methanol were not
available for deep freezing purposes. The maximum COP values of the cycle were
reported as 1.6 for cooling and 2.63 for heating purposes.
A prototype of a regenerative two-bed adsorption cooling cycle was designed and
experimentally tested by Critoph et al. in 2000 [57]. Activated carbon ammonia as
the working pair and water as the heat transfer fluid were selected. The results of this
study showed that with a maximum bed temperature of 140
o
C, condensation temper-
21
ature of 32
o
C, evaporation temperature of 7.5
o
C and half cycle time as 620s the ob-
tained COP value of the cycle was 0.44 and specific cooling power was approximate-
ly 0.18kW/kg.
A new design for two-stage adsorption cooling systems with two adsorbent beds was
analyzed by Saha et al. in 2001 [58]. In this two-stage system, the pressurization (or
depressurization) was completed in two progressive stages as the name implies. To
achieve this aim, each adsorbent bed was divided into two parts, and as one part was
being heated, the other part was being cooled. More details can be found in the relat-
ed article. Silica gel water was selected as the adsorbent refrigerant (working)
pair for the maximum bed temperature range of 40-75
o
C. With a maximum tempera-
ture of 55
o
C and condensation temperature of 30
o
C, the COP of this cycle was re-
ported as 0.36.
Wang in 2001 investigated the effects of heat and mass recovery on the simple cycle
with two adsorbent beds and activated carbon methanol pair [38]. The experimental
results stated with the addition of heat recovery, the COP values increased by ap-
proximately 25%. Implementing mass recovery before heat recovery added 10% to
the COP value of a simple cycle with heat recovery. Results also showed that with
the mass recovery after heat recovery, COP reached 0.5 for a maximum bed tempera-
ture of 100
o
C, condensation temperature of 24
o
C, evaporation temperature of 10
o
C
and heat capacity ratio of 1.85. Heat capacity ratio is defined as the ratio of the beds
design to inherent heat capacities (R), i.e., ratio of heat capacities of adsorbent beds
shell and heat transfer fluid to heat capacity of adsorbent inside the bed. The results
were also verified by theoretical analyses.
A prototype of a solar powered simple adsorption cooling system was built in Nige-
ria by Anyanwu and Ezekwe in 2003 [59]. This designed system was tested with
activated carbon methanol pair in November and December 1995. Therefore, the
collector efficiencies were low and the obtained COP of the cooling cycle was
around 0.31-0.35.
22
Additionally, experimental studies were done by Wang et al. in 2003 to investigate
the effects of heat and mass recovery [60, 61]. In these studies, two beds with acti-
vated carbon methanol pair were used with maximum bed temperatures lower than
120
o
C and evaporation temperature at 10
o
C. According to the results, implementing
mass recovery to the adsorption cycle increased the cycles cooling power up to 11%
where the maximum COP was 0.115. The maximum COP value increased to 0.125
with the implementation of both heat and mass recovery processes. Additionally,
results showed that heat recovery processes can decrease the total required heat
transfer to drive the cycle by 20-30%.
Another design which was a simple adsorption cooling cycle with a binary working
medium was proposed by Wang and Zhu in 2004 [62]. Instead of a cycle which
worked with a zeolite water pair, the authors suggested using a zeolite ammonia
water working media. Therefore, the cycle would work with a significantly larger
pressure swing than a cycle with a zeolite water pair. This cycle is actually a com-
bination of adsorption and absorption cooling cycles. The test results showed that the
COP increased considerably, but COP values were dependent on the concentration of
ammonia in the working medium. Additionally, this new system was more respon-
sive to temperature changes. A maximum COP value of 0.48 was obtained when the
maximum bed temperature was 120
o
C.
A simple adsorption cooling cycle was also developed by Chang et al. in 2007 using
a silica gel water pair [63]. With a novel design of the adsorbent bed, a COP of
0.53 was obtained for the test conditions of maximum bed temperature of 80
o
C, con-
densation temperature of 30
o
C and evaporation temperature of 14
o
C. The results also
showed that the COP value of the cycle decreased as the maximum bed and evapora-
tion temperatures decreased and condensation temperature increased. Within this
study, effects of mass flow rate in the condenser and cycle time on the cycles COP
values were also investigated.
A solar powered adsorption cooling system was developed by Zhai and Wang in
2009 for the green house in Shanghai Jiao Tong University [64]. They provided 10
o
C
23
of evaporation temperature using a maximum bed temperature of 85
o
C and conden-
sation temperature of 32
o
C for a silica gel water pair. They designed this simple
adsorption system with evacuated tube collectors and wet cooling towers, and exper-
imental COP values were within 0.29-0.32. Economic analyses were also done for
this system and the corresponding results showed that the payback period for this
solar-powered adsorption cooling system coupled with heating and hot water appli-
cations was approximately 3 years.
An adsorption cooling cycle with activated carbon HFC 134a pair was experimen-
tally investigated by Banker et al. in 2010 [65]. They worked with maximum bed
temperatures between 73-93
o
C, condensation temperatures between 28-32
o
C and
evaporation temperatures between 5-18
o
C. Their study included steady and transient
results and their aim was to cool down some electronic components. However, they
could only test their model up to 5W cooling loads and COP values obtained in this
study were below 0.04 for all investigated configurations.
Grisel et al. in 2010 constructed a heat recovery adsorption cooling system with a
two-bed silica gel water pair and analyzed its performance experimentally [66].
They tested the system at maximum temperatures between 73-91
o
C, condensation
temperatures between 22-43
o
C and evaporation temperatures between 6-20
o
C. When
heat recovery was considered, the maximum COP value of the cycle was approxi-
mately 0.85.
2.2.3 Previously Developed Numerical Modeling
a. Simple adsorption cooling cycle
A description of the operating principles and processes of the simple adsorption cy-
cle are given in Section 2.2.1.a and a thermodynamic model of the simple adsorption
cycle is presented in this section. The thermodynamic model presented here was pre-
viously developed by Baker and Kaftanolu [35, 36, 67].
24
The main goal of the model is to predict the coefficient of performance of the adsorp-
tion cycle (COP
ads
). In order to predict COP
ads
, heat transfer values for each process
have to be calculated. The required heat transfer to the adsorbent bed for process
12 (Q
F,1-2
) can be found using the 1
st
Law of Thermodynamics as,

,1 2
d d d d d
F ads shell HTF ref
Q U U U U
= + + + (2.1)
( ) ( ) ( ) ( )
,1 2
d d d d d
F
ads shell HTF ref
Q m u m u m u m u
= + + + (2.2)

,1 2
d d d d d d
F ads ads shell shell HTF HTF ref ref ref ref
Q m u m u m u m u u m
= + + + + (2.3)
Note that d d d 0
ads shell HTF
m m m = = = and the mass of vapor refrigerant is neglected
(therefore, no change in adsorbed refrigerant, d 0
ref
m = ). Assuming constant specific
heats,

,1 2
,1 2
d
d d d d
ref F
shell shell HTF HTF
F ads ads ref
ads ads ads ads
m Q
m c m c
q c T c T c T
m m c m
+
= = + + (2.4)
Using the definition of adsorption capacity (X) and defining the ratio of the adsorbent
beds design to inherent heat capacities (R) as ( )( )
1
shell shell HTF HTF ads ads
R m c m c m c

= +
where
( )
shell shell HTF HTF
m c m c + is the design heat capacity and
( )
ads ads
m c is the inher-
ent heat capacity,

( )
,1 2 1
d 1 d d
F ads ref
q R c T X c T
= + + (2.5)
Therefore, the required heat transfer for process 12 can be found by integrating
Equation (2.5). The required heat transfer for process 23 can be calculated using a
similar approach to that given in Equations (2.1) through (2.5), except the refrigerant
vapor leaves the adsorbent bed (i.e., d 0
ref
m = ) for process 23. Therefore,
25

,2 3
d d d d d d
F v ads shell HTF ref
Q H U U U U
+ = + + + (2.6)
( ) ( ) ( ) ( ) ( )
,2 3
d d d d d d
F
v ads shell HTF ref
Q m h m u m u m u m u
+ = + + + (2.7)

,2 3
d d d d d d d
F v v ads ads shell shell HTF HTF ref ref ref ref
Q h m m u m u m u m u u m
+ = + + + + (2.8)
Similar to Equations (2.4) and (2.5) with d d d 0
ads shell HTF
m m m = = = and
d d 0
ref v
m m = = ,

( )
,2 3
d d d d d d
F ads ads shell shell HTF HTF ref ref ref v ref
Q m u m u m u m u u h m
= + + + + (2.9)
Defining a constant heat of adsorption as
ads ads v
h u h A = ,

( )
,2 3
d 1 d d d
F ads ref ads
q R c T X c T h X
= + + +A (2.10)
The required heat transfer for process 23 can be found by integrating Equation
(2.10). Modeling process 34 is similar to modeling process 12, except that the
adsorbent bed cools down for process 34. Therefore, the heat transfer from the
adsorbent bed to the environment can be calculated similar to Equation (2.5) as,

( )
,3 4 3
d 1 d d
bed ads ref
q R c T X c T
= + + (2.11)
The heat transfer from the adsorbent bed for process 41 can be calculated using
Equation (2.10) with refrigerant vapor entering the adsorbent bed at the evaporation
temperature (T
evap
).
Heat transfers from the condenser and the evaporator per unit mass of adsorbent can
also be calculated, respectively, as,
( ) ( ) ( )
v liq 3 1
h h
cond o o
q T T X X ( =

(2.12)
26

( ) ( ) ( )
v liq 3 1
h h
P evap o
q T T X X
(
=

(2.13)
where X
3
X
1
is described as the adsorption capacity swing of the cycle, AX
cycle
. The
energy performance of the adsorption cycle is defined as the coefficient of perfor-
mance of the adsorption cycle (COP
ads
), which is also one of the main parameters
examined in this study.

,1 2 ,2 3
P P
ads
F F F
q q
COP
q q q

= =
+
(2.14)
b. Heat recovery
The mathematical model is constructed for heat recovery cycle which is described in
Section 2.2.1.b using equations similar to Equations (2.1)-(2.14). Heat transfer that is
being transferred between the two adsorbent beds during heat recovery (q
rec
) is calcu-
lated considering the corresponding sorption regime (sorption regimes are described
in Section 2.2.1.b) for each simulated case. The COP
ads
of heat recovery cycle is cal-
culated similarly with COP
ads
of the simple cycle given in Equation (2.14), where
F,HRec F,Simple rec
q q q = . Details are given in [35, 67].
c. Mass recovery
The mass recovery model does not require any special mathematical formulation
beyond the simple adsorption cycles relations. The two beds at the end of their ad-
sorption/desorption processes are brought into mechanical equilibrium until the two
adsorbent beds have the same pressure. As suggested by Qu et al. [37], for ideal mass
recovery process and for simplicity, the adsorption capacity (X) changes of the two
adsorbent beds over the mass recovery process are also assumed to be identical. De-
tails are given by Qu et al. [37].
27
d. Thermal wave adsorption cooling cycle
Thermal wave adsorption cooling cycle is introduced briefly in Section 2.2.1.d. In
this section, a thermodynamic model of thermal wave adsorption cooling cycle that
was developed by Baker and Kaftanolu [35] and re-evaluated by Taylan et al. [44]
is given in detail for comprehensiveness of the present study.
To explore thermodynamic limits to and trends for COP
ads
for the thermal wave ad-
sorption cooling cycle, the following assumptions are made. Each bed is spatially
(but not temporally) isobaric and the only heat exchange is with the heat transfer
fluid (HTF). An ideal square thermal wave divides each adsorption bed into a back
region behind the wave and a front region in front of the wave. Each region is iso-
thermal and isosteric with the back region at T
b
T
HTF,in
and X
b
, and the front region
at T
f
T
HTF,out
and X
f
. Sorption at the thermal wave occurs at the average wave tem-
perature, T
w
= (T
b
+ T
f
)/2. Each bed has a normalized length of 1 and the position of
the moving thermal wave is 0 y
w
1. The adsorbent and adsorbed refrigerant mass-
es for the back and front regions are given mathematically in Table 2.1.
Table 2.1. Adsorbent and adsorbed refrigerant masses for the back and front regions
of the bed [44].
Back region at T
b
& X
b
Front region at T
f
& X
f

Mass of adsorbent (m
ads
)
w ads
y m

( ) 1
w ads
y m

Mass of adsorbed refrigerant (m
ref
)
w ads b
y m X

( ) 1
w ads f
y m X

A half-cycle is completed when the waves reach the right sides of the beds in Figure
2.6. The cycle is completed by reversing the direction of the HTF flow via a reversi-
ble pump, thereby reversing the direction of the thermal wave. At state 1, y
w,1
= 0,
T
b,1
= T
f,1
= T
o
, P
1
= P
evap
and all valves are closed. A heat capacity ratio, R, analo-
gous to that introduced in Equation (2.5) is defined as the ratio of the design to the
28
inherent heat capacities [38, 67]. The model is developed for the left bed undergoing
a complete cycle in Figure 2.6 and initially uniformly at the minimum bed tempera-
ture (T
o
).
Process 12: The heat transfer fluid starts to circulate at T
hot
through the bed initiat-
ing the thermal wave. At the wave the refrigerant desorbs at T
w
causing the vapor
pressure to increase, which in turn causes refrigerant vapor to be re-adsorbed in the
front and back regions. For simplicity, the heat of adsorption released in the back
region is assumed to be transferred to the thermal wave, such that T
b
T
hot
and dT
b
=
0. In reality, the re-adsorption of refrigerant vapor occurring behind the wave likely
causes the temperature behind the wave to vary spatially and temporally. However,
modeling these temperature variation increases the complexity and computational
time of the model significantly. Since one of the goals of the present work is to use
fast models to perform a large number of parametric studies, as a first case approxi-
mation the effect of these spatial and temporal variations is assumed small and is
neglected. Conversely, in the front region the heat of adsorption causes the tempera-
ture to increase (dT
f
> 0). To find T
f
and the vapor pressure inside the bed, a control
volume at an unspecified position, y
s
, in front of the wave is considered (y
w
< y
s
< 1).
The control volume is spatially both isothermal (T
s
= T
f
) and isosteric (X
s
= X
f
). Some
net amount of refrigerant vapor entering the control volume is adsorbed (dm
v,in
=
dm
ref,s
), while the rest passes through the control volume. An energy balance can be
obtained similar to Equation (2.5),

( ) ( ) ( )
, , , , , ,
d 1 d d
v in v in ads s ads s ref s ref s
h m R m u m u = + + (2.15)
Rearranging the terms in Equation (2.15) yields,

( ) d 1 d d d
v,in ref,s ads,s ads,s ref,s ref,s ref,s ref,s
h m R m u m u u m = + + + (2.16)
Using the relations in Table 2.1 and noting that dm
ads,s
= 0,
29

( ) d 1 d d d
v,in ads,s s ads,s ads,s s ads,s ref,s ads,s ref,s s
h m X R m u X m u m u X = + + + (2.17)
Canceling out m
ads,s
in all the terms,

( ) d 1 d d d
v,in s ads,s s ref,s ref,s s
h X R u X u u X = + + + (2.18)
Rearranging the terms in Equation (2.18) yields,
( ) ( )
1 d d d 0
ads, f f ref, f ref, f v,in f
R u X u u h X + + + = (2.19)
Using constant specific heats and adding d
v, f f
h X to the right hand side of Equation
(2.19),
( ) ( )
1 d d d 0
ads f f ref f ref, f v, f v, f v,in f
R c T X c T u h h h X + + + + = (2.20)
Applying constant specific heats and rearranging terms yield,
( ) ( ) ( )
1 d d d 0
ads f ref f ref, f v, f f p f in f
R c X c T u h X c T T X ( + + + + =

(2.21)
A more detailed and complete presentation of Equation (2.21) that explicitly shows
the dependence of u
ref
and h
v
on temperature is,

( )
( ) ( ) ( )
ref v
1 d u h d 0
ads f ref f f f p f in f
R c X c T T T c T T X ( + + + + = (

(2.22)
Equation (2.22) is independent of y
s
and is valid for the entire front part of the bed.
Assuming the desorbed refrigerant vapor enters the control volume at T
in
= T
w
and
constant heat of adsorption [68], Equation (2.22) becomes,

( ) ( )
1 d d 0
ads f ref f ads p f w f
R c X c T h c T T X
( ( + + + A + =

(2.23)
30
From Equation (2.23) it is possible to find T
f
implicitly for a specified vapor pres-
sure. To find the position of the thermal wave, y
w
, conservation of refrigerant mass
can be applied to the adsorbent bed.

ref ref,b ref, f
m m m = + (2.24)
Since no refrigerant leaves the bed during the process 12, dm
ref
= 0. Therefore, m
ref

in Equation (2.24) is constant during this process. Using the relations in Table 2.1,

( ) 1
ads min w ads b w ads f
m X y m X y m X = + (2.25)
where X
b
= X(T
hot
, P
v
), X
f
= X(T
f
, P
v
) and X
min
= X(T
o
, P
evap
). Therefore, the position
of the thermal wave, y
w
, can be found using,

( ) ( )
( ) ( )
X , X ,
X , X ,
o evap f v min f
w
b f hot v f v
T P T P X X
y
X X T P T P

= =

(2.26)
An energy balance on the entire bed for dm
HTF,flow
yields,

( ) ( ) ( ) ( )
d 1 d d
HTF,in HTF,out HTF, flow ads,b ads,b ads, f ads, f ref,b ref,b ref, f ref, f
h h m R m u m u m u m u = + + + +
(2.27)
Note that T
b
=T
hot
and T
hot
is fixed, therefore du
ads,b
= du
ref,b
= 0. Conservation of mass
necessitates dm
ads,b
+ dm
ads,f
= 0 and dm
ref,b
+ dm
ref,f
= 0 with dm
ads,b
0, dm
ads,f
0,
dm
ref,f
0 and dm
ref,b
0. Therefore, Equation (2.27) becomes,

( ) ( ) ( )
( )
d 1 d d
d d
HTF,in HTF,out HTF, flow ads,b ads, f ads,b ads, f ads, f
ref,b ref, f ref,b ref, f ref, f
h h m R u u m m u
u u m m u
(
= + +

+ +
(2.28)
Using constant specific heats,
31

( ) ( ) ( )
( )
d 1 d d
d d
HTF in out HTF, flow ads b f ads,b ads, f ads f
ref b f ref,b ref, f ref f
c T T m R c T T m m c T
c T T m m c T
(
= + +

+ +
(2.29)
Using Table 2.1, Equation (2.29) becomes,

( ) ( ) ( ) ( )
( ) ( )
( )
d 1 d 1 d
d d
1 d
HTF b f HTF, flow ads ads b f w w f
ads ref b f b w ads ref b f w b
w ads ref f f
c T T m R m c T T y y T
m c T T X y m c T T y X
y m c X T
(
= + +

+ +
+
(2.30)
Rearranging terms in Equation (2.30) yields,

( ) ( ) ( )
( )
( )
{ }
,
d
1 1 d
1 d d
HTF flow
HTF hot f w ads ref f f
ads
hot f ads ref b w ref w b
m
c T T y c R c X T
m
T T c R c X y c y X
( = + +

( + + + +

(2.31)
For fixed P
v
, T
f
and y
w
, Equation (2.31) can be solved for dm
HTF,flow
. Process 12
ends when P
v
= P
cond
.
Process 23: The valve between the bed and the condenser is opened, and the wave
continues to progress through the bed. The desorbed refrigerant leaves the bed and is
condensed in the condenser, resulting in a temporally isobaric process inside the bed
at P
cond
. An energy balance on the entire bed yields,

( ) ( ) ( )
( )
, ,
d d 1 d
d
HTF,in HTF,out HTF, flow v out v out ads,b ads,b ads, f ads, f
ref,b ref,b ref, f ref, f
h h m h m R m u m u
m u m u
= + +
+ +
(2.32)
Since all desorbed refrigerant leaves the bed, dT
f
= dT
b
= 0. From conservation of
mass, dm
ads,b
= -dm
ads,f
= m
ads
dy
w
and dm
v,out
= -dm
ref,b
dm
ref,f
, where m
v,out
is the
mass of refrigerant vapor exiting the bed. Assuming the refrigerant leaves the bed at
T
w
and using Table 2.1 and a similar methodology as to obtain Equation (2.31) from
Equation (2.27),
32

( )
( ) ( ) ( )
( ) ( )
d
1
d
HTF hot f HTF, flow
ads b f ads hot f
ads w
p b hot w f f w
c T T m
h X X R c T T
m y
c X T T X T T
= A + +
(
+

(2.33)
Integrating Equation (2.33) yields,

( )
( ) ( ) ( )
( ) ( )
2 2 3
2 2 2
2
2 2 2
1
1
HTF hot f, HTF, flow,
ads b, f, ads hot f,
w, ads
p b, hot w f, f, w
c T T m
h X X R c T T
y m
c X T T X T T
= A + +
(
+

(2.34)
Equation (2.34) fixes m
HTF,flow,2-3
. Process 23 ends when y
w
= 1 and the bed is uni-
formly at T
hot
and P
cond
.
Process 34: The valve between the bed and the condenser is closed, the direction
of the heat transfer fluid flow reversed, and the heat transfer fluid now enters the bed
at T
o
. The process is essentially the reverse of process 12 and described by a set of
equations parallel to Equations (2.15)-(2.31). Equations (2.23) and (2.31) is valid
with T
w
= T
evap
and T
b
= T
o
and Equation (2.26) becomes,

( ) ( )
( ) ( )
X , X ,
X , X ,
hot cond f v max f
w
b f o v f v
T P T P X X
y
X X T P T P

= =

(2.35)
Process 41: The valve between the bed and the evaporator is opened, and the wave
continues to progress through the bed. The process is essentially the reverse of pro-
cess 23 with the exception that evaporated refrigerant enters the bed at T
evap
. Equa-
tions parallel to Equations (2.32) and (2.34) describe this process as,

( )
( ) ( ) ( )
( ) ( )
4 4 1
4 4 4
4
4 4 4
1
1
HTF o f, HTF, flow,
ads b, f, ads o f,
w, ads
p b, o evap f, f, evap
c T T m
h X X R c T T
y m
c X T T X T T
= A + +
(
+

(2.36)
33
q
htr
and q
clr
: For the processes 123, the total mass of heat transfer fluid that pass-
es through the bed is,

2
1 3 2 3
1
d
HTF, flow, HTF, flow HTF, flow,
ads ads ads
m m m
m m m

= +
}
(2.37)
Similarly, for processes 341,

4
3 1 4 1
3
d
HTF, flow, HTF, flow HTF, flow,
ads ads ads
m m m
m m m

= +
}
(2.38)
Significantly, m
HTF,flow,1-3
< m
HTF,flow,3-1
since the adsorption capacities are different
for processes 12 and 34, resulting in different sensible loads [42]. This differ-
ence in m
HTF,flow
yields two possible thermal wave cycle models: without bypass and
with bypass.
In the thermal wave cycle without bypass model, all m
HTF,flow,3-1
passes through both
the heater and cooler. Therefore, the total heat transfers per unit mass of adsorbent to
the heater and from the cooler are,

( ) ( )
4
4 1 4
3
d
ads htr
hot f HTF, flow HTF, flow, hot f,
HTF
m q
T T m m T T
c

= +
}
(2.39)

( ) ( )
( )( )
2
2 3 2
1
3 1 1 3
d
ads clr
f o HTF, flow HTF, flow, f, o
HTF
HTF, flow, HTF, flow, hot o
m q
T T m m T T
c
m m T T

= +
+
}
(2.40)
In the thermal wave cycle with bypass model, m
htf,flow,1-3
passes through the heater
while m
htf,flow,3-1
m
htf,flow,1-3
bypasses the heater and the heated bed. Therefore, only
m
htf,flow,1-3
needs to be heated to T
hot
resulting in,
34

( ) ( )( )
4
1 3 3 4 4
3
d
ads htr
hot f HTF, flow HTF, flow, HTF, flow, hot f,
HTF
m q
T T m m m T T
c

= +
}
(2.41)
Since all m
htf,flow,3-1
needs to be cooled to T
o
,

( ) ( )
( )( )
2
2 3 2
1
3 1 1 3 4
d
ads clr
f o HTF, flow HTF, flow, f, o
HTF
HTF, flow, HTF, flow, f, o
m q
T T m m T T
c
m m T T

= +
+
}
(2.42)
Note that while q
clr
given in Equations (2.40) and (2.42) is not needed for COP
ads

calculations, it is used in a global energy balance to check the 1
st
Law consistency of
the overall model.
e. Reversible cycle
In the all adsorption cycles described in Sections 2.2.3.a through 2.2.3.d, irreversibil-
ities occur due to heat transfers across finite temperature differences and throttling.
In order to have the maximum possible thermodynamic performance, a reversible
cycle is modeled. One completely reversible cycle between the three thermal reser-
voirs shown in Figure 2.4 is a combination of a Carnot refrigerator operating be-
tween T
o
and T
evap
that is driven by a Carnot power cycle operating between T
hot
and
T
o
. Therefore, COP
ads
of the reversible cycle is given in [69] as,

1
1
o
hot
ads,rev
o
evap
T
T
COP
T
T
(2.43)
The reversible cycle has the maximum theoretically possible COP
ads
and COP
ads
of
reversible cycle does not depend on the components or the processes of the cycles.
For this reason, the reversible cycle is a base case with which to compare all the ad-
sorption cycles described in Sections 2.2.3.a through 2.2.3.d.
35
2.2.4 Previous Theoretical Studies
Theoretical models of adsorption systems with their mathematical explanations can
be widely found in literature. For brevity, only a number of the relevant studies are
chronologically summarized in this study. As before, more information can be found
in the references provided for each study.
Douss et al. in 1988 developed a numerical model for simple and heat and mass re-
covery cycles [70]. Heat pumps with zeolite NaX water pair were considered and
analyzed in this theoretical study. The numerical model was supported by some ex-
perimental study. According to the results, the simple cycle could reach COP of 1.38
whereas heat and mass recovery cycle could reach 1.56. The results also stated that
condensation and evaporation pressures had the largest effect on the COP values
among bed temperatures and other components design parameters (such as heat
transfer coefficients).
Effects of operating conditions on a heat recovery adsorption heat pump were inves-
tigated by Zheng et al in 1995 [71]. In this study, activated carbon ammonia pair
was selected and two adsorbent beds were used. Conditions like maximum bed tem-
perature and ambient temperature were analyzed. The results showed that increasing
maximum bed temperature from 180 to 250
o
C increased the COP by 20%. Addition-
ally, as the ambient temperature (which ideally is equal to the condensation tempera-
ture if a dry cooling tower is used) increased, COP decreased.
Another theoretical study was conducted by Zheng et al. in 1995 [72]. They analyzed
a two-bed adsorption heat pump with heat regeneration was analyzed, and an activat-
ed carbon ammonia pair was used. Effects of several parameters related to cycle
time and heat transfer coefficients on the cycle performance were investigated, but
the related part of this study is the effect of the dead mass within the adsorbent bed.
According to the results, increasing the adsorbent bed shells mass (i.e., heat capaci-
ty) decreased the COP and cooling capacity of the cycle whereas increasing the mass
36
of heat transfer fluid within the adsorbent bed increases COP slightly, but decreases
cooling capacity considerably.
Another model was developed for simple and heat recovery cycle by Teng et al. in
1997 [73]. In this study, effects of the parameters (coefficients and exponent) in Du-
binin Astakhov (D-A) equation on cycle performance were investigated. More in-
formation on D-A equation can be found in Section 2.3. Additionally, effects of max-
imum bed temperature, evaporation temperature and heat capacity ratio were ana-
lyzed. As predicted, the adsorption cycle with regeneration gave higher COP values
than the simple cycle. Additionally, COP values of the adsorption heat pump with
activated carbon methanol pair increased by increasing the maximum bed tempera-
ture (varied between 90 140
o
C) and evaporation temperature (varied between -20
15
o
C) or decreasing heat capacity ratio (varied between 0 50). The maximum COP
value reported was approximately 1.5.
A thermal wave adsorption heat pump model with only heat transfer equations (i.e.,
excluding mass transfer) was developed by Sun et al. in 1997 [74]. Zeolite NaX
ammonia pair was selected with constant heat of adsorption value for analyses. The
model considered two-dimensional heat transfer within the adsorbent bed: axial for
heat transfer fluid only (excluding axial heat transfer for adsorbent) and radial heat
transfer. This two-dimensional model was solved using ordinary differential equa-
tions (one dimensional) with a number of assumptions. As a result, the authors re-
ported a maximum COP of 0.87 for the following condition: maximum bed tempera-
ture at 260
o
C, condensation temperature at 40
o
C, evaporation temperature at 5
o
C and
ambient temperature at 20
o
C.
Pons and Poyelle in 1999 investigated the heat recovery and thermal wave adsorption
cycles numerically [75]. These cycles were coupled with mass recovery, and two
types of working pair were selected: activated carbon methanol and zeolite NaX
water. Additionally, effects of cooling tower type on the system performance were
investigated. The COP values of the cycle obtained in this study are given in Table
2.2.
37
Table 2.2. COP comparison of adsorption systems with different configurations [75].
Heat and Mass Recovery Cycle Thermal Wave Cycle with Mass
Recovery
Zeolite NaX
Water
Activated Car-
bon Methanol
Zeolite NaX
Water
Activated Car-
bon Methanol
Cooling
Tower
Dry Wet Dry Wet Dry Wet Dry Wet
Cooling 0.63 0.68 0.52 0.55 0.70 0.92 - 0.61
Ice
Making
- - 0.33 0.42 - - - 0.31
Heat
Pump
1.6 1.7 1.45 1.6 1.7 2.0 - -

According to the results, the thermal wave cycle gave higher COP values than heat
recovery cycle (except for ice-making) and wet cooling tower was advantageous over
dry cooling tower in terms of COP. Note that ice making results for zeolite NaX
water pair are not available since water cannot be run below 0
o
C (which is the freez-
ing temperature of water) in adsorption cooling cycles. When the authors compared
these results with the previous studies, they claimed that adding mass recovery en-
hanced the COP values.
A numerical model for thermal wave adsorption cycle with zeolite NaX water pair
was developed by Sward et al. in 2000 [76]. The effects of maximum bed tempera-
ture (varied between 100
o
C and 120
o
C) and condensation temperature (varied be-
tween 30
o
C and 40
o
C) on COP values were analyzed. For the selected base case
where maximum bed temperature was 120
o
C, condensation temperature was 30
o
C
and evaporation temperature was 5
o
C, a COP of 1.24 was reported. Additionally, as
the maximum bed temperature increased and as the condensation temperature de-
creased, COP values of the cycle was increased.
38
Wang et al. in 2000 [77] predicted the performance of solar powered ice maker cou-
pled with water heater through energy analysis. Activated carbon methanol pair
was selected as working pair and simulations were performed for a complete year.
The results showed that the maximum COP of the refrigeration cycle was 0.51 for
the period from October to December which resulted according to the simulated op-
erating conditions. The simulation results were higher than the experimental results
in which the COP was 0.386 in two days in December 1998.
Effects of heat transfer fluid in a thermal wave adsorption cooling cycle with two
beds were investigated numerically through thermodynamic models by Pons and
Szarzynski in 2000 [78]. Zeolite water pair was used, and heat transfer fluid with
constant density and specific heat was compared with heat transfer fluid with tem-
perature dependent density and specific heat. According to the simulation results,
assuming constant density did not affect the performance of the cycle whereas results
with constant specific heat assumption over-predicted the cycle performance.
Chua et al. in 2001 investigated the differences in performance of the adsorption
cooling cycles as the number of adsorbent beds was varied [51]. When two, four and
six adsorbent beds were compared in a heat recovery adsorption cooling cycle, heat
recovered increased by 70% switching from two beds to four beds and by 40%
switching from four beds to six beds. Silica gel water pair was considered in this
study with the base case as a maximum bed temperature at 85.7
o
C and condensation
temperature at 31
o
C.
Two and three-bed adsorption cycles with heat recovery and silica gel water pair
were compared by Saha et al. in 2003 [79]. The results stated that with three adsor-
bent beds it was possible to increase the recovered heat transfer by 35% compared to
system with two adsorbent beds. The parametric results showed that the maximum
COP value (0.43) was achieved when the maximum bed temperature was between
80
o
C and 90
o
C if the condensation temperature was 30
o
C. COP increased with in-
creasing maximum bed temperature and with decreasing condensation temperature.
39
A theroretical study of the previously developed two-stage adsorption cooling cycle
[58] was accomplished using a theoretical model of the system by Alam et al. in
2004 [80]. Silica gel water pair was selected as in the experimental study, and ef-
fects of cycle time, maximum bed and condensation temperature and mass of adsor-
bent on the system performance were investigated. According to the results, COP
values increased, achieved a peak value and then decreased as the maximum bed
temperature increased. Cooling capacity increased with decreasing cycle time and
increasing mass of adsorbent, while with longer cycle times better COP values could
be attained. Finally, as the condensation temperature decreased, both COP and cool-
ing capacity increased.
Wang et al. in 2005 developed an analytical model for a two-bed adsorption cooling
cycle with silica gel water pair [81]. Effects of heat and mass recovery and operat-
ing temperatures on the system performance were investigated. The highest COP was
obtained when the maximum bed temperature was 80-85
o
C for the investigated 65-
85
o
C range. However, mass recovery enhancement was more effective at low maxi-
mum bed temperatures. When the condensation temperature was 20
o
C, the maximum
predicted COP of the cycle was 0.65.
Another theoretical model for a solar-powered regenerative adsorption heat pump
was developed by Lambert in 2007 [82]. The author suggested using activated car-
bon ammonia pair with CaCl
2
additions. Therefore, within the system physical ad-
sorption could be supported by chemical adsorption. As a result, the adsorption ca-
pacity of the pair increased by 35% on average. Detailed design parameters and cor-
responding mathematical model are presented in the provided reference. According
to the results, a maximum COP of 1.604 was reported. Additionally, different types
of solar collectors (single or double-glazed flat plate with or without coating, evacu-
ated tube and parabolic concentrator) were compared under certain conditions and
the highest solar thermal efficiency was obtained from evacuated tube collector while
parabolic concentrator had second highest solar thermal efficiency.
40
Liu and Leong claimed that the condensation pressure is not constant as time pro-
gresses for the process 23 due to heat transfer limitations which were not included
in previous studies. A numerical two-dimensional heat and mass transfer model that
accounted for heat transfer limitations was developed by Liu and Leong in 2008 for a
transient simple adsorption cooling cycle [83]. Contradictory to the authors claim,
the simulations results showed that the pressure in the condenser does not change
significantly for the simple adsorption cycle and COP increases as mass flow rate
within the condenser increases (in this model, systems with wet cooling tower was
simulated).
A new design called an adsorption tube was developed by Wang and Zhang in 2009
[84]. An adsorption tube is actually an adsorption cooling (or heating) system that
consists of an adsorbent bed, a condenser and an evaporator in a single tubular hous-
ing. In order to verify the accuracy of this new design, the authors analyzed the sys-
tem with a numerical model. Although the system was small in size compared with
the previous studies, the COP of the cycle for silica gel water pair was approxi-
mately 0.5. The authors predicted that one of the main concerns would arise in the
manufacturing process since a thin walled housing with low heat capacity is neces-
sary in order to decrease the cycle time.
A solar-powered two-bed adsorption cooling cycle with heat and mass recovery was
modeled for silica gel water pair by Luo et al. in 2010 [85]. The COP values were
predicted using energy balances with a maximum bed temperature range of 55-90
o
C.
The results predicted that a COP of the system higher than 0.25 can be achieved un-
der fair solar radiation and with an evacuated tube collector. Results also stated that
the evacuated tube collector always has higher thermal efficiencies than the flat plate
collector for the investigated cases.
2.3 Adsorption Capacity Models
Adsorption capacity (X) models of adsorbent refrigerant (working) pairs were de-
veloped and presented in the literature. More information on working pairs and the
41
properties of adsorbents and refrigerants can also be found in these review papers
[24, 32, 33, 86, 87]. Only the final governing equations for these models are present-
ed here for completeness and brevity.
In general, adsorption capacity (X) is characterized by temperature (T) and pressure
(P) and given in form of Dubinin Astakhov (D-A) equations [88]. There are two
main D-A equations, Equations (2.44) and (2.45), to describe the adsorption capacity
(X) as a function of temperature and pressure.

( )
( )
sat
X , exp 1
T
n
o
T
T P X k
P
(
| |
(
=
|
|
(
\ .

(2.44)

( )
( )
sat
X , exp ln
P
n
o
P
T P X D T
T
(
| |
(
=
|
|
(
\ .

(2.45)
Working pairs which have been widely investigated and are well-established are cho-
sen for the present study. Investigated pairs and their corresponding coefficients for
D-A equations are given in Table 2.3.
Table 2.3. Coefficients for D-A equations [89].
Adsorbent Refrigerant Pair X
o
k D n
Zeolite NaX Water (Z1) [90] - - - -
Zeolite X13 Water (ZW) [24, 91] 0.261 5.36 - 1.73
Silica Gel Water (SG) [24, 92] 0.350 - 610
-6
1.70
Activated Carbon Ammonia (CA) [24, 93] 0.290 3.57 - 1.38
Activated Carbon Methanol (CM) [24, 93] 0.450 13.38 - 1.50

42
One exception to using the D-A equation is for the zeolite NaX water (Z1) pair. Its
adsorption capacity model is presented in a different form and details are given in
[90, 94]. The adsorption capacity model for zeolite NaX water (Z1) pair is de-
scribed with Equations (2.46) through (2.49).

2, 3, 4,
1, 2 3
, 1, 2
j j j
j j
a a a
ns a for j
T T T
= + + + = (2.46)

3 1 2
ns d ns ns = (2.47)

,
exp , 1, 2, 3
j
j o j
E
b b for j
T
| |
= =
|
\ .
(2.48)

3
1
1
j j
j
j
P
X ns b
b P
=
=
+
(2.49)
The constants in Equations (2.46)-(2.49) are given in Table 2.4.
Table 2.4. Coefficients for adsorption capacity model of zeolite NaX water pair
[90, 94].
j a
1
a
2
a
3
a
4
b
o
E d
1 0.070 -119.9 63690 -8450000 1.50810
-10
7726
2 -0.687 775.7 -254200 27750000 5.40710
-10
6075 0.267
3 - - - - 1.70810
-10
5392

In the adsorption cycle models, there are several other constants that are specific to
the adsorbent refrigerant pair. These constants are heat of adsorption (Ah
ads
) [87,
95], specific heats of adsorbent (c
ads
) and refrigerant vapor (c
p
) and the enthalpy of
vaporization (h
fg
) and are given in Table 2.5.
43
Table 2.5. Constants for adsorption cycle models [89].
Adsorbent
Refrigerant Pair
Ah
ads

[kJ/kg]
c
ads

[kJ/kg
.
K]
c
liq

[kJ/kg
.
K]
c
p

[kJ/kg
.
K]
h
fg

[kJ/kg]
Activated Carbon
Ammonia (CA) [24, 93]
1900 0.930 4.84 2.13 1225.03
Activated Carbon
Methanol (CM) [24, 93]
1900 0.930 1.405 1.405 1200
Silica Gel
Water (SG) [24, 92]
2500 1.000 4.2 1.8644 2476.9
Zeolite NaX
Water (Z1) [90]
3200 0.836 4.2 1.8644 2476.9
Zeolite X13
Water (ZW) [24, 91]
3750 0.836 4.2 1.8644 2476.9

The working pairs given in Table 2.3 and the zeolite NaX water pair are embedded
in the adsorption cycle models stated in Sections 2.2.3 and 3.1.
2.3.1 Previous Studies on Working Pair Comparison
A number of studies on working pair comparison are presented in the literature. In
this section, some of the results from those papers are given chronologically. As be-
fore, more explanations about each paper can be found in the provided references.
Performance of simple and heat recovery adsorption cycles are simulated by Wang et
al. in 1997 for granular and activated carbon fiber methanol pairs [96]. According
to the results, COP was increased by almost 20% with the addition of heat recovery
with two adsorbent beds, and COP values of systems with activated carbon fiber
were 15% higher on average than systems with granular activated carbon.
44
Effects of adding salts such as CaCl
2
, LiBr, LiCl, MgCl
2
into the well-established
adsorbents such as silica gel, activated carbon, zeolite on the performance of simple
cycle were analyzed by Restuccia et al. in 2004 [97]. The experimental results were
also supported by a theoretical study, and as a result, addition of salts improved the
COP values of the simple adsorption cycle by 30% for cooling and 60% for heating
cycles when zeolite 4A water and silica gel water pairs were considered at the
same conditions.
The previous approach was also experimentally analyzed by Wang et al. in 2004
[98]. They suggested adding activated carbon to CaCl
2
to form a new hybrid adsor-
bent for ice-making in adsorption refrigeration cycle where ammonia was selected as
refrigerant. As the experimental results showed, addition of activated carbon to
CaCl
2
increased the adsorption capacity until a maximum value. Therefore, there was
an optimum addition level of activated carbon to obtain the best performance among
the investigated conditions. The authors stated that adding activated carbon increased
the mass transfer inside the adsorbent bed.
Cui et al. in 2005 compared several adsorbent refrigerant pairs, with adsorbents as
zeolite 13X, silica gel, activated carbon and composite adsorbents developed by the
authors (NA and NB) and refrigerants as water and ethanol [99]. Adsorption and
cooling capacities of these working pairs were tested experimentally where cooling
capacity was defined as the adsorption swing of the adsorption capacity times the
latent heat of adsorbent. The result of this study stated that NA water pair had al-
most 2.5 times higher adsorption capacity as zeolite 13X water pair had. Similarly,
NB ethanol pair had 3 times higher adsorption capacity than activated carbon
ethanol pair. These results are also valid for cooling capacities.
A numerical study of adsorption refrigeration cycles with two types of activated car-
bon fiber methanol pairs were done by Hamamoto et al. in 2006 [100]. Specific
cooling effects and coefficient of performances for simple adsorption cycle with two
adsorbent beds were compared within two activated carbon fiber methanol pairs
and also with silica gel methanol pair. The results stated that activated carbon fiber
45
methanol and silica gel methanol pairs have distinct ranges of operating condi-
tions ranges where COP values of one pair is higher than others.
Saha et al. in 2006 investigated the adsorption performances of silica gel water,
activated carbon ethanol and activated carbon HFC 134a [101]. The conditions
were selected as evaporation temperature at 7
o
C, condensation temperature at 30
o
C
and the maximum bed temperature range between 50-90
o
C. Results are given in
terms of isentropic efficiency and uptake efficiency which is a function of adsorption
capacity. According to the results, isentropic efficiency decreases as the maximum
bed temperature increases.
Different activated carbon types with ethanol as refrigerant were experimentally test-
ed by El-Sharkawy et al. in 2008 [102]. As a result, correlations for adsorption ca-
pacities for these pairs were obtained using D-A equations. Clapeyron diagrams were
also obtained for the investigated working pairs.
Loh et al. in 2009 analyzed six different pairs theoretically for simple adsorption
cycles [103]. These pairs included two different activated carbon fiber ethanol,
chemviron R134a, fluka R134a and maxsorbll R134a. According to this studys
results, as the maximum bed temperature increases, the COP values of the simple
cycle also increases. For the investigated two activated carbon fiber ethanol pairs,
maximum COP values were 0.56 and 0.68 where the maximum bed temperature was
85
o
C. For the silica gel water pair, the maximum COP value at 85
o
C maximum bed
temperature was 0.68. Results also showed that activated carbon ethanol pair has
the highest specific cooling power while silica gel water pair gives the highest
cooling effect per unit volume of adsorbent.
2.4 TRNSYS Software
TRNSYS is a FORTRAN-based simulation software that is used throughout the pre-
sent study. The name TRNSYS stands for the Transient Energy System Simulation
tool [104]. This software is commercially available since 1975 and developed by the
46
members of Solar Energy Laboratory at the University of Wisconsin Madison.
However, the foundation of the software was a joint project between the University
of Wisconsin Madison and the University of Colorado [105].
In general, TRNSYS is a tool that is used to simulate the performance of transient
energy systems, especially thermal systems. TRNSYS includes several built-in com-
ponents which can be used for constructing the desired systems with ease. These
components generally include algebraic or differential equations which are also
solved numerically within the software. One of the strengths of this software is that
other companies (such as TESS Thermal Energy System Specialists, Inc. [106])
also develop other components which are not available in the default library of the
software as it is purchased. Additionally, users can develop their own components
with the provided source codes. Using this software, the users are able to;
- display the simulation results as the simulation continues so that users do not
have to wait for the whole simulation to be completed,
- provide inputs to the software with simple text files and get the outputs as
formatted spreadsheets,
- form complex user graphical interfaces which allow users to enter inputs and
get the predefined outputs,
- finalize the model with executable files
- call some other software, such as MATLAB, Fluent and Excel, within TRN-
SYS.
More details on this software can be found on these websites [105-107].
2.5 Objectives of the Present Study
Models for the ideal simple adsorption cycle, the ideal heat recovery adsorption cycle
with two spatially isothermal beds and ideal thermal wave model exist in the litera-
ture with corresponding analyses. The present study builds on these aforementioned
existing works by:
47
- Implementing the existing adsorption cycle models into a TRNSYS-
compatible environment via MATLAB
- Extending the thermal wave model to include adiabatic and isothermal mass
recovery with or without by-pass line
- Modeling the following commonly-used adsorbent refrigerant (working)
pairs using MATLAB
o Zeolite NaX Water
o Zeolite X13 Water
o Silica Gel Water
o Activated Carbon Ammonia
o Activated Carbon Methanol
- Constructing a solar thermal system model within TRNSYS that can be used
both for steady and seasonal-transient simulations
- Integrating three (two flat plate and one evacuated tube) commercial collector
models with the solar thermal system
- Developing a normalized seasonal model where solar and loss fractions, col-
lector area and mass of adsorbent can be analyzed
- Linking the adsorption cycle models with working pair models and the solar
thermal system to construct solar-thermal-powered adsorption cooling system
- Running steady and seasonal-transient simulations
- Investigating basic trends in the cycle and system performances as the follow-
ing parameters are varied
o Maximum bed temperature
o Condensation temperature
o Evaporation temperature
o Heat capacity ratio
o Excess bed temperature
48
CHAPTER 3

MODELS
The models presented in this chapter are extensions of models previously published
or under revision by Taylan, Baker and Kaftanolu [44, 89, 108] and build on the
models presented in Chapter 2.
3.1 Modifications on Thermal Wave Adsorption Cooling Cycle
Model
In this section, definitions and thermodynamic models of analyzed adsorption cycles
are given. In all models, the following assumptions are considered.
- Constant specific heat and heat of adsorption
- Refrigerant vapor as an ideal gas
- Mass of refrigerant vapor in the adsorbent bed is negligible relative to that in
the liquid and adsorbed phrases.
- The modeled adsorbent beds are assumed to be spatially isobaric.
- Enthalpy change in throttling is neglected.
- Specific heat of adsorbed refrigerant is assumed to be equal to the specific
heat of refrigerant vapor to be consistent with the 1
st
Law of Thermodynamics
[90].
Three thermal wave cycles are considered within this study: (1) no mass recovery
(NMR); (2) adiabatic mass recovery (AMR); and, (3) isothermal mass recovery
(IMR). For the no mass recovery thermal wave cycle (NMR), the vapor line and
mass recovery valve connecting the two beds in Figure 2.6 do not exist. For the mass
recovery cycles, mass recovery occurs at the end of each half cycle when the hot,
49
high-pressure bed is at T
hot
and P
cond
, and the cold, low-pressure bed is at T
o
and P
evap

< P
cond
. During the mass recovery process, the two beds are isolated from the con-
denser and evaporator, and the mass recovery valve is opened. Refrigerant vapor
flows from the hot, high-pressure bed to the cold, low-pressure bed until the two
reach the same pressure. Adiabatic mass recovery (AMR) and isothermal mass re-
covery (IMR) thermal wave cycles represent two limiting cases. For AMR, each bed
undergoes an adiabatic mass recovery process, which is equivalent to the heat trans-
fer fluid (HTF) pump being turned off. For IMR, each bed undergoes an isothermal
mass recovery process, which is equivalent to the HTF pump being left on.
Clapeyron diagrams for both back and front regions of NMR, AMR and IMR are
given in Figure 3.1 and Figure 3.2, respectively.
3.1.1 Thermal Wave Adsorption Cooling Cycle with Adiabatic Mass Recovery
(AMR)
For the mass recovery cycles, at the end of the half cycles (states 1 and 3) the valves
connecting the beds to the condenser and evaporator are closed, and the mass recov-
ery valve connecting the beds is opened to allow refrigerant vapor to flow from the
high pressure to the low pressure bed. For AMR, the beds are assumed adiabatic
throughout the mass recovery process, and therefore, there is no heat exchange with
the HTF. For the back region, as shown in Figure 3.1, the processes 12 and 34 in
NMR are divided into the two processes 1a1a1
+
2 and 3a3a3
+
4. Simi-
larly, for the front region the corresponding processes and states for AMR are
1a1a2 and 3a3a4 as shown in Figure 3.2.
For the processes 1a1 and 3a3 (i.e., the mass recovery processes), assuming that
all the refrigerant desorbed in the high-pressure bed will be adsorbed by the low-
pressure bed as suggested by [37],

m m min n max n
X X X X X X A = = = A (3.1)
50

Figure 3.1. Clapeyron diagram for back region of thermal wave adsorption cycles
(Notation for States: # = NMR, a# = AMR, i# = IMR, s# = Simple Cycle, while a#
+

and i#
+
occur immediately after corresponding a# and i#) [44].

Figure 3.2. Clapeyron diagram for front region of thermal wave adsorption cycles
(Notation for States: # = NMR, a# = AMR, i# = IMR, s# = Simple Cycle) [44].
where subscripts m and n denote the low-pressure and high-pressure beds, respec-
tively.
Defining the final common pressure as P
a
= P
a1
= P
a3
, and referring to Figure 3.1 and
Figure 3.2, yields,
51

( ) ( ) ( ) ( )
1 2
X , X , X , X ,
a a o evap hot cond a a
T P T P T P T P = (3.2)
The exact path for the mass recovery process is defined by assuming the refrigerant
vapor entering the low-pressure bed is at the same temperature as the high-pressure
bed,
( ) ( ) ( )
d 1 d d
v v ads ads ref ref
h m R m u u m = + + (3.3)
Note that the masses of adsorbent in both beds are constant, i.e., dm
ads
= 0. Using the
definition of adsorption capacity, X,

( ) d 1 d d d
v ads m ads ads ads m ref ads ref m
h m X R m u m X u m u X = + + + (3.4)
Canceling out m
ads
in all terms and rearranging Equation (3.4),
( ) ( ) ( )
ref v
1 d d u h d 0
ads m ref m n m
R u X u T T X + + + = (

(3.5)
Note that the refrigerant vapor enters the low-pressure bed at T
n
, while the adsorbent
bed itself is at T
m
. Adding
( )
v
h d
m m
T X to Equation (3.5) yields,
( ) ( ) ( ) ( ) ( )
ref v v v
1 d d u h h h d 0
ads m ref m m m n m
R u X u T T T T X + + + + = (

(3.6)
Applying constant specific heats to Equation (3.6) gives,
( ) ( ) 1 d d d 0
ads m m ref m ads p m n m
R c T X c T h c T T X ( + + + A + =

(3.7)
Arranging terms in Equation (3.7) yields,

( ) ( ) 1 d d 0
ads p m m ads p m n m
R c c X T h c T T X ( ( + + + A + =

(3.8)
52
A parallel set of equations similar to Equations (3.3)-(3.8) can be derived for the
high-pressure bed, except for the high-pressure bed the refrigerant vapor leaves the
adsorbent bed at T
n
,

( ) 1 d d 0
ads p n n ads n
c R c X T h X ( + + + A =

(3.9)
The processes a1a1
+
2, a3a3
+
4, a1a2 and a3a4 in Figure 3.1 and Figure
3.2 are identical to processes 12 and 34 of NMR, except they start at T
a1
and P
a
,
and T
a3
and P
a
. Therefore, the model for these processes is parallel to Equations
(2.15)-(2.42).
3.1.2 Thermal Wave Adsorption Cooling Cycle with Isothermal Mass
Recovery (IMR)
Modeling the mass recovery processes for IMR is similar to that with AMR except
the adsorbent beds are assumed isothermal through heat exchange with the HTF.
Therefore, the model is parallel to Equations (2.15)-(2.42) except the adsorbent beds
are kept both spatially and temporally isothermal during the mass recovery process.
Unlike AMR, to keep both beds at constant temperatures requires heat transfer to
occur to the hot bed and from the cold bed during the mass recovery process. The
heat transfer per unit mass of adsorbent can be calculated for the low-pressure bed
similar to Equation (3.8) as,

( )
i1
1 1
1
d
bed, i
ads p o hot
ads
q
h c T T X
m

( = A +
}
(3.10)
since the refrigerant enters the low-pressure bed at T
hot
= T
3
. Similar to Equations
(3.9) and (3.10), for the high-pressure bed,

i3
3 3
3
d
bed, i
ads
ads
q
h X
m

= A
}
(3.11)
53
The processes i1i1
+
2, i3i3
+
4, i1i2 and i3i4 in Figure 3.1 and Figure 3.2
are parallel to processes 12 and 34 of NMR, except the processes start at T
i1
= T
1

and P
i
, and T
i3
= T
3
and P
i
, respectively.
3.2 Solar Thermal System Model
The solar thermal system consists of a solar thermal collector, a heat exchanger, a
circulation pump and a control unit. Details of each component are given in the fol-
lowing subsections. The solar thermal system is modeled in TRNSYS, and built-in
types of TRNSYS are used to model these components. The schematic presentation
of the solar thermal system is given in Figure 3.3.

Figure 3.3. Solar thermal system and its integration with the adsorption cycle [89].
A TRNSYS screen capture of the solar thermal system is also given in Figure 3.4. In
Figure 3.4, the solid lines represent the flow loop of the working fluid and the dashed
lines represent the information flow between the TRNSYS components.
54

Figure 3.4. TRNSYS screen capture of the solar thermal system.
3.2.1 Solar Thermal Collectors
The solar thermal collector is one of the essential components of the solar thermal
powered adsorption cooling system. Two flat plate collectors (coded as FPL and FP)
and one evacuated tube collector (coded as ET) are modeled using built-in TRNSYS
types. The collectors corresponding performance parameters are obtained from
commercial solar thermal collectors and these parameters are given in the following
two subsections. The manufacturers of these collectors are not provided so as not to
commercialize the present study.
a. Flat plate solar collectors
TRNSYS built-in component Type 1c is used to model the flat plate collectors. This
component is based on the second order efficiency formula which is given in Equa-
tion (3.12). The form of this equation is based on ASHRAE 93-77 and 96-1980
standards and most of the collector manufacturers use this form of the efficiency re-
lation to get solar certification for their collectors [109].

( )
2
0 1 2
mean amb mean amb

coll
T T T T
d d d
G G

= (3.12)
55
q
coll
is the thermal efficiency of the solar collector, T
mean
(
o
C) is the mean tempera-
ture of the inlet and outlet temperatures of the solar collector, T
amb
(
o
C) is the ambient
temperature, G (W/m
2
) is the solar radiation on the solar collector and d
0
, d
1
and d
2

are performance coefficients that are given for a specific solar collector. The perfor-
mance coefficients that are used in modeling are given in Table 3.1.
Table 3.1. Performance coefficients for the flat plate solar collectors.
Collector Code FPL FP
Conversion Coefficient
( )
0
d
0.780 0.751
Loss Coefficient
( ) 2 1
W
m K
d

3.591 4.999
Loss Coefficient
( ) 2 2 2
W
m K
d

0.0199 0.000

The other parameters that are required to model flat plate solar collectors are given in
Table 3.2.
Table 3.2. Other necessary parameters for the flat plate solar collectors.
Collector Code FPL FP
Collector Area
( )
2
m 1.912 1.620
Specific Heat of Working Fluid
( )
kJ
kg K
4.190 4.190
Tested Flow Rate
( )
2
kg
h m
74.791 53.500

56
The final parameters required to model the flat plate collectors using TRNSYS Type
1c are the Incidence Angle Modifiers (IAM). The Incidence Angle Modifiers define
the fraction of radiation incident on the collector cover that reaches the absorber of
the collector. The modifier is one when the collector cover transmits all solar radia-
tion, and it is zero when the collector cover does not transmit any solar radiation.
According to ASHRAE standards, incidence angle modifiers include both direct and
diffuse light modifiers [109]. The IAM values that are used in modeled flat plate col-
lectors are given in Table 3.3.
Table 3.3. Incidence angle modifiers for the flat plate collectors.
Collector Code FPL
Incidence Angle 0
o
50
o
90
o

IAM 1.000 0.865 0.000
Collector Code FP
Incidence Angle 0
o
20
o
30
o
40
o
50
o
60
o
70
o
90
o

IAM 1.000 0.983 0.958 0.918 0.851 0.732 0.484 0.000

b. Evacuated tube solar collector
TRNSYS built-in component Type 71 is used to model an evacuated tube collector.
This component also uses the second order efficiency formula as given in Equation
(3.12). One of the main differences between the flat plat collector and evacuated tube
collector models is the Incidence Angle Modifiers (IAM). Evacuated tube collectors
require biaxial (both longitudinal and transverse) IAM values while flat plate collec-
tors require only one dimensional IAM values. The parameters that are used to model
the commercial evacuated tube solar collector in TRNSYS Type 71 are given in Ta-
ble 3.4.
57
Table 3.4. Parameters for the evacuated tube solar collector.
Collector Code ET
Conversion Coefficient
( )
0
d
0.825
Loss Coefficient
( ) 2 1
W
m K
d

1.190
Loss Coefficient
( ) 2 2 2
W
m K
d

0.009
Collector Areas
( )
2
m 2.000
( )
kJ
kg K
4.190
Tested Flow Rate
( )
2
kg
h m
3.0 (TRNSYS Default Value)

3.2.2 Heat Exchanger
TRNSYS built-in component Type 92 is used to model the heat exchanger shown in
Figure 3.3. The main purpose of using this heat exchanger in the solar thermal sys-
tem is to provide the heat transfer that is required to drive the adsorption cooling cy-
cle (q
F
). The hot working fluid that exits the solar collector provides the required q
F

as shown in Figure 3.3. TRNSYS Type 92 also has an internal on-off controller
whose control signal can be sent externally, and this signal is sent by the control unit
as shown in Figure 3.3. The parameters of the modeled heat exchanger are given in
Table 3.5. Note that in order to analyze the limits of solar thermal system, the maxi-
mum capacity of the heat exchanger is chosen as a very large value and all the losses
in heat exchanger are neglected.
58
Table 3.5. Parameters for the heat exchanger.
Maximum Capacity
( )
kJ
h
100,000
( )
kJ
kg K
4.190
Loss Coefficient
( )
kJ
h K
0.000
Heat Exchanger Efficiency (-) 1.000

3.2.3 Circulation Pump
TRNSYS built-in component Type 3d is used to model the circulation pump in Fig-
ure 3.3. Although this component has the capability to vary the flow rate of the work-
ing fluid that is circulating in the loop shown in Figure 3.3, in this solar thermal sys-
tem the flow rate is kept constant. As in the heat exchanger, all the losses in the cir-
culation pump are neglected. The parameters that are used to model the circulation
pump are given in Table 3.6.
Table 3.6. Parameters for the circulation pump.
Maximum Flow Rate
( )
kg
h
100
( )
kJ
kg K
4.190
Maximum Power
( )
kJ
h
60
Loss Coefficient (-) 0

59
3.2.4 Control Unit
TRNSYS built-in component Type 2b is used to model the control unit shown in
Figure 3.3. This control unit is added to the solar thermal system to fix the tempera-
ture at which the heat transfer that is required to drive the adsorption cooling cycle
(q
F
) occurs since in the modeled ideal adsorption cooling system this temperature is
kept constant at the maximum bed temperature (T
hot
).
The operating principle of this control unit is as follows: If the outlet temperature of
the solar collector is higher than the initially set T
hot
, the control unit turns on the heat
exchanger and the outlet temperature of the heat exchanger becomes the set T
hot
. On
the other hand, if the outlet temperature of the solar collector is lower than the initial-
ly set T
hot
, the control unit turns off the heat exchanger and the adsorption cycle does
not operate until the outlet temperature of the solar collector reaches T
hot
.
3.2.5 Weather Data
In addition to the units that are described in the preceding subsections, another TRN-
SYS built-in component Type 109-TMY is used to obtain the meteorological data
(for instance, ambient temperature and solar radiation). TMY stands for Typical Me-
teorological Year, and all the meteorological data that are in TMY format can be read
by TRNSYS Type 109-TMY.
3.2.6 Integration of Solar Thermal System with Adsorption Cycle Models
As mentioned in the preceding sections, adsorption cooling cycles are modeled in
MATLAB and the solar thermal system is modeled in TRNSYS. The TRNSYS
screen of the integration of the solar thermal system and adsorption cooling cycle
models is given in Figure 3.5. In Figure 3.5, Macro is the solar thermal system as
shown in Figure 3.4, Type 155 is the component that is used to run MATLAB com-
mands, Equa includes the cells to enter the inputs to the system, Equa-2 has the for-
mulation to calculate the system efficiency, Type 65d shows the requested outputs to
60
show on the screen and Type 25c writes all the outputs to a specified Microsoft Excel
file. A sample TRNSYS deck (input) file for a solar system integrated thermal wave
cycle with evacuated tube collector is given in Appendix A and TRNSYS Studio
Report that shows the relations between the TRNSYS components is given in Ap-
pendix B for the same system.

Figure 3.5. Integration of solar thermal system with adsorption cycle in TRNSYS.
As mentioned in Appendix B, the inputs for the integrated TRNSYS model are,
- Adsorbent refrigerant pair
- Minimum adsorbent bed temperature (T
o
)
- Maximum adsorbent bed temperature (T
hot
)
- Condensation temperature (T
cond
)
- Evaporation temperature (T
evap
)
- Ratio of beds design to inherent heat capacities (R) defined as
( )( )
1
shell shell HTF HTF ads ads
R m c m c m c

= +
The coefficient of performance of the adsorption cooling cycle (COP
ads
) and the effi-
ciency of the solar thermal collector (q
coll
) are obtained as outputs of the integrated
TRNSYS model, and the overall system efficiency (COP
sys
) is calculated by multi-
plying the coefficient of performance of adsorption cooling cycle (COP
ads
) by the
efficiency of solar thermal collector (q
coll
).
61
3.3 Normalized Model
A new normalized model is developed for post-processing of the simulation results.
The normalized model is used to investigate how the size of the cooling system,
amount of storage, and coincidence between the solar-supplied cooling and cooling
demand affects the systems seasonal energy performance. In all cases this normal-
ized model stresses simplicity over detail to quickly identify the most promising di-
rections for more detailed research.
Several normalized transient parameters are defined at each time step (t
i
). A normal-
ized cooling load (q
load,N
) is defined as being proportional to the temperature differ-
ence between T
amb
and a reference temperature (T
rfrc
).

( )
( )
( )
amb
load,N
T
q
Max
i rfrc
i
amb rfrc
t T
t =
T T
(3.13)
Here Max(T
amb
T
rfrc
) corresponds to the time interval during the simulation with the
highest T
amb
, and therefore maximum cooling load, and 0 1
load,N
q s s . A normalized
cooling capacity (q
clg,N
) is defined as,

( )
( ) ( )
( )
F ads
clg,N
q COP
q
Max
i i
i
F ads
t t
t = S
q COP
(3.14)
where S is a user defined size of the cooling system and 0 q
clg,N
S. Additionally,
the required heat transfer is calculated using
( )
F HTF p,HTF hex,in hex,out
q =m c T T where
HTF is the heat transfer fluid of the solar thermal system as shown in Figure 3.3. By
definition, one unit of normalized load is equal to one unit of normalized cooling
capacity and S > 1 corresponds to the common case where the maximum output of
the cooling system is larger than the maximum cooling load, which allows for stor-
age. At each time step, the transient parameter q
Match,N
is used to quantify the differ-
ences in the magnitudes of the normalized load and normalized cooling capacity as,
62

( ) ( ) ( )
Match, N clg,N load,N
q q q
i i i
t = t t (3.15)
Thus q
Match,N
< 0 indicates the case where backup cooling or storage is needed to
meet the cooling load while q
Match,N
> 0 indicates an opportunity for storage. Three
normalized parameters are defined to investigate how the amount of storage affects
broad trends in the systems seasonal energy performance. To keep the model as
simple as possible, only the storage magnitude and not type (hot/cold storage, build-
ing thermal mass, etc.) is specified. The user inputted normalized storage magnitude,
q
Storage,max
, is defined such that one unit of normalized storage (q
Storage,N
) is equal to
one unit of normalized cooling load (and capacity). At the start of each simulation
the storage is assumed empty; i.e., q
Storage,N
(t=0) = 0. A normalized loss, q
Loss,N
, quan-
tifies the solar-supplied cooling lost to the environment when q
clg,N
> q
load,N
and the
storage is full. The normalized backup power, q
Backup,N
, quantifies the backup cooling
power required when q
clg,N
< q
load,N
and q
Storage,N
= 0.
For each simulation, solar and loss fractions are defined as additional seasonal ener-
gy performance parameters. The solar fraction (f) is the normalized load met by solar
energy.

( )
( )
clg,N
load,N
q
q
i
clg,tot
i
load,tot i
i
t
q
f = =
q t
(3.16)
The loss fraction, l, is defined as,

( )
( )
Loss,N
load,N
q
q
i
Loss,tot i
load,tot i
i
t
q
l = =
q t
(3.17)
Similar to the transient and seasonal energy performance parameters, two normalized
size parameters for collector area and mass of adsorbent are also defined. A normal-
ized collector area (A
coll,N
) is defined as,
63

( )
( ) ( )
clg,N
N sys
q
G COP
i
coll,N
i i
t
A
t t
= (3.18)
where
( ) ( )
N
G G
i i rfrc
t = t G is the normalized solar radiation and G
rfrc
is a reference
solar radiation level. Therefore, one unit of normalized collector area supplies one
unit of normalized cooling capacity when both G
N
and COP
sys
are one. As a normal-
ized area, A
coll,N
can be used to compare the relative required collector area to meet
the cooling load for different cases. A
coll,N
is constant in each individual case but var-
ies among the investigated cases.
Similar to the normalized collector area, a normalized mass of adsorbent (m
ads,N
) is
defined. As an intermittent cooling cycle, the required m
ads,N
decreases with decreas-
ing cycle time. For simplicity in these analyses all cycle times are assumed equal.
The required m
ads,N
is inversely proportional to maximum change in adsorption ca-
pacity of the adsorbent over the half cycle (AX = X
max
X
min
) during the simulation
relative to that for a base case (AX
base
).

( )
-1
Max
ads,N
base
X
m =
X
A (
(
A

(3.19)

64
CHAPTER 4

ANALYSES AND VERIFICATION
The information presented in this chapter is an extension of that previously published
or under revision by Taylan, Baker and Kaftanolu [44, 89, 108].
4.1 Analyses
A number of steady and seasonal-transient simulations are run using the modeled
system described in detail in Section 2.2.3 and Chapter 3 to investigate general ener-
gy performance trends while several parameters are varied. Seasonal transient simu-
lations are run with 15-minute time intervals over an entire summer from June 1 to
September 30 using hourly weather data for Antalya, a city on Turkeys Mediterra-
nean coast, using the typical meteorological year format as described in Section
3.2.5. In all simulations, liquid water is used as heat transfer fluid (HTF), and its
properties are taken from [21, 110].
As pointed out in previous studies [44, 108], system performance depends on several
parameters in addition to the working pair. In all simulations, the energy perfor-
mance characteristics of either a flat plate or evacuated tube non-tracking solar col-
lector located on the roof of the Mechanical Engineering Department at the Middle
East Technical University are used. The energy performance characteristics for these
collectors obtained from their manufacturers are given in Section 3.2.1. Another in-
vestigated parameter is the condensation temperature (T
cond
) (or cooling tower type).
The effect of using a dry rather than wet cooling tower due to concerns over water
consumption is investigated through the condensation temperature. The condensation
temperature is set equal to either ambient temperature (which corresponds to using
an ideal dry cooling tower) or wet bulb temperature (which corresponds to using an
65
ideal wet cooling tower). Evaporation temperature (T
evap
) is also investigated, and it
is fixed by the application of the cooling system (i.e., the required temperature for
cooling). The investigated values of the heat capacity ratio of the adsorbent bed (R)
correspond to the theoretical limit (R = 0), state-of-the-art design (R = 3) and typical
design (R = 10) [38]. Minimum and maximum bed temperatures are also investigat-
ed. It is theoretically possible to cool down the adsorbent bed to the condensation
temperature. However, in reality due to the heat capacities of the system and some
other design restrictions, the possible lowest temperature of the adsorbent bed (T
o
)
may be larger than the condensation temperature (T
cond
). In order to consider this
condition, AT
excess
is defined as AT
excess
= T
o
T
cond
. Maximum bed temperature (T
hot
)
is dependent on the working pair. Analyzed values for these parameters with the in-
vestigated working pairs are summarized in Table 4.1.
In order to determine T
hot
ranges for a specific working pair, two main criteria are
applied to the adsorption capacity models of the pairs given in Table 4.1. One criteri-
on is based on a minimum adsorption capacity swing for a half cycle,
( )
Min 0.03
cycle max min
X X X A = = , since cooling capacity is directly proportional to
AX, and operating the cycle with low AX values will not be advantageous. The other
criterion is applied to X
min
. X
min
values that are lower than 0.03 are disregarded since
adsorption capacities (X) approach to some value (in this study this value is chosen
as 0.03) asymptotically (i.e., zero adsorption capacity is physically not possible) and
very small X values require large increases in temperature. Additionally, in order to
eliminate inconsistencies embedded in the adsorption capacity models, X
max
values
that are higher than
( ) ( ) ( )
o o
sat
X , X 25 C, P 15 C
o evap evap
T P T = = are eliminated, as
among the investigated conditions given in Table 4.1, the maximum X value can be
obtained with
o
25 C
o
T = and
( )
o
sat
P 15 C
evap evap
P T = = . All these criteria yield the
T
hot
ranges given in Table 4.2.

66
Table 4.1. Investigated conditions.
Parameter Analyzed Values or Condition
Adsorption Cycle Type Reversible
Ideal Simple
Heat Recovery with Two Spatially Iso-
thermal Beds
Thermal Wave with No Mass Recovery
(with by-pass line)
Thermal Wave with No Mass Recovery
(without by-pass line)
Thermal Wave with Adiabatic Mass Re-
covery (with by-pass line)
Thermal Wave with Adiabatic Mass Re-
covery (without by-pass line)
Thermal Wave with Isothermal Mass
Recovery (with by-pass line)
Thermal Wave with Isothermal Mass
Recovery (without by-pass line)
Adsorbent Refrigerant (Working) Pair Zeolite NaX Water (Z1)
Zeolite X13 Water (ZW)
Silica Gel Water (SG)
Activated Carbon Ammonia (CA)
Activated Carbon Methanol (CM)
Cooling Tower Type Dry, Wet
Condensation Temperature, T
cond
(
o
C) 25, 30, 35
Evaporator Temperature, T
evap
(
o
C) 5, 10, 15
Heat Capacity Ratio, R 0, 3, 10
Excess Bed Temperature, AT
excess
(
o
C) 0, 5, 10
Maximum Bed Temperature Range, T
hot
(
o
C) 50 250

67
Table 4.2. Analyzed maximum bed temperature ranges for each pair.
Adsorbent Refrigerant Pair T
hot
range (
o
C) T
hot
values (
o
C)
Zeolite NaX Water (Z1) 90 190 90, 120, 150, 180, 190
Zeolite X13 Water (ZW) 80 160 80, 100, 120, 150, 160
Silica Gel Water (SG) 80 100 80, 90, 100
Activated Carbon Ammonia (CA) 80 150 80, 100, 120, 150
Activated Carbon Methanol (CM) 80 100 80, 90, 100

For the thermal wave cycles with adiabatic (AMR) or isothermal (IMR) mass recov-
ery, a finite difference method is applied to solve Equations (3.8)-(3.11). For AMR,
mass recovery processes are divided into a series of steps with fixed AX increments
and decrements. At each step, the bed temperature for the low-pressure bed (T
m
) is
found using Equation (3.8). A similar discretization scheme is used for IMR except
since the temperatures are known at each step, the corresponding pressures (P
m
and
P
n
) can be found iteratively from X = X(T,P). Steps are continued to be taken until P
n

P
m
< 0. Subsequently, linear interpolations between the last two steps are done for
both beds until a convergence criterion (|P
n
P
m
| < c) is reached. When the conver-
gence criterion is met, states a1 and a3 for AMR and i1 and i3 IMR are fixed.
To interpret the possible effects of the parameters given in Table 4.1 and Table 4.2
on each systems energy performance, cases are compared based on COP
ads
, COP
sys

and COP
sys,clg
. COP
ads
is given in Equation (4.1).

( ) ( ) ( )
v l
,3 3 ,3 3
h h
max min evap cond
evap
ads
htr bed i htr bed i
X X T T
q
COP
q q q q
(

= =
+ +
(4.1)
Note that for the simple and heat recovery cycles q
htr
= q
F
. For the simple and heat
recovery cycles, and for the thermal wave cycles with no mass (NMR) and adiabatic
mass recovery (AMR) q
bed,3i3
= 0. COP
ads
of the reversible cycle can be calculated
using the information in Section 2.2.3.e and as given in Equation (2.43).
68
The systems overall coefficient of performance can be calculated as,

sys ads coll
COP COP = (4.2)
where the definition of collector efficiency (q
coll
) is given in Section 3.2.1.
A coefficient of performance of the integrated system (COP
sys
) for steady simula-
tions as given in Equation (4.2) and cooling capacity weighted COP
sys
(COP
sys,clg
) for
seasonal-transient simulations are defined as performance parameters. The definition
of COP
sys,clg
is given in Equation (4.3).

( ) ( )
( )
sys clg,N
clg,N
COP q
q
i i
i
sys,clg
i
i
t t
COP =
t
(

(4.3)
To further compare the investigated parameters given in Table 4.1, COP
sys,clg
ratio
(r
sys
) is introduced such that,

( )
( ) ( )
( )
sys,clg,i sys,clg,base
sys,i
sys,clg,base
COP , COP ,
r ,
COP ,
hot hot
hot
hot
T WP T WP
T WP
T WP
= (4.4)
where i represent any case, WP is the abbreviation of the working pair, and base case
is selected with simple cycle, dry cooling tower, T
evap
= 10
o
C, R = 10 and AT
excess
=
0
o
C.
The constants in the normalized model described in Section 3.3 are as follows. To fix
A
coll,N
in the normalized model, G
rfrc
is chosen as 1,000 W/m
2
which is close to the
maximum solar radiation level of Antalya, Turkey. Similarly, to calculate the cooling
load (q
load,N
), T
rfrc
= 21
o
C is selected. For m
ads,N
, AX
base
is calculated assuming T
hot
=
150
o
C, T
o
= T
cond
= 30
o
C, P
cond
= P
sat
(T
cond
), and P
evap
= P
sat
(T
evap
) where P
sat
indi-
cates the saturation pressure as a function of temperature.
69
4.2 Error Quantification
The adsorption cycle models described in detail in Section 3.1 are checked for con-
sistency using the 1
st
Law of Thermodynamics. For this reason, an energy balance is
performed for the overall system. For the simple and heat recovery cycles, the system
boundary is drawn around the adsorbent beds, condenser and evaporator.

htr clr bed
htr
q q q
Error
q

= (4.5)
Similarly, another error is defined for the thermal wave cycles. The system boundary
for the thermal wave cycles also includes the heater and the cooler as partially shown
in Figure 2.6. The error for thermal wave cycles is given in Equation (4.6).

( )
( )
,1 1 ,3 3 2
,3 3
htr clr bed i bed i p max min w, evap
htr bed i
q q q q c X X T T
Error
q q
+ +
=
+
(4.6)
Note that for the thermal wave cycles with no (NMR) and adiabatic (AMR) mass
recovery cycles, q
bed,1i1
= q
bed,3i3
= 0 in Equation (4.6).
Additional consistency checking for models is established through the 2
nd
Law of
Thermodynamics. The adsorption models are empirical curve fits to experimental
data and may not necessarily obey the 2
nd
Law of Thermodynamics. Therefore, the
COP
ads
values of the investigated pairs are compared with the reversible COP
ads
, and
cases with COP
ads
higher than the reversible COP
ads
are eliminated.
The assumption of constant specific heats is also investigated by comparing the
model results with model results in which variable specific heat values are used. This
verification is specifically done for working pairs with water as refrigerant since wa-
ter properties are well established. Therefore, the adsorption cycle model for NMR
(both with and without by-pass line) is modified and several cases are run using vari-
able specific heats. Sample steady simulation durations and corresponding percent-
70
age errors using Equation (4.6) for NMR and by-pass line using T
hot
= 150
o
C, T
cond
=
30
o
C, T
evap
= 10
o
C, AT
excess
= 0
o
C, R = 10 are given in Table 4.3. In Table 4.3, the
interval size refers to the number of steps between states 1 and 2 (and states 3 and 4)
as shown in Figure 3.1 and Figure 3.2, which is used in the numerical solution of
integrodifferential equations in Sections 2.2.3 and 3.1.
Table 4.3. Simulation durations and 1
st
Law error comparison of constant and varia-
ble specific heats (NMR cycle with by-pass line, T
hot
= 150
o
C, T
cond
= 30
o
C, T
evap
=
10
o
C, AT
excess
= 0
o
C and R = 10).
Interval Size Constant Specific Heat Variable Specific Heat
Duration (s) Error (%) Duration (s) Error (%)
1000 5.185 0.17221 32.832 0.14605
5000 19.224 0.03442 147.265 0.01106
10000 35.990 0.01721 321.018 0.00580
50000 221.594 0.00344 1487.501 0.01929

The comparison results in Table 4.3 show that the simulation times increase signifi-
cantly as the model switches from constant to variable specific heats while the per-
centage error that is described in Equation (4.6) decreases slightly. Therefore, con-
stant specific heats are assumed to reduce simulation times and allow more simula-
tions to be run. Additionally, using constant specific heat gives higher COP
ads
values
than variable specific heat, which is consistent with findings of Pons and Szarzynski
[78]. For all analyzed cases, as the interval size of the finite-difference scheme is
halved, the error is almost halved as can be interpreted from Table 4.3, suggesting
that this error originates from the finite-difference method.
71
4.3 Solar Thermal Collector Model Verification
The solar thermal system is modeled as described in Section 3.2 and one of the main
components of the solar thermal system is the solar thermal collectors. In order to
verify the accuracy of the collector models, outlet temperatures of the modeled col-
lectors are compared with experimental results. The experiments were conducted in
the Mechanical Engineering Department at the Middle East Technical University in
2008 [111]. The experimental parameters such as the inlet temperature of the collec-
tor, solar radiation, ambient temperature and the flow rate of the working fluid are
manually entered to the collector model, and the outlet temperatures of the actual and
modeled collectors are compared.
In Table 4.4 and Table 4.5, verification results for commercially available flat plate
solar collectors are given (FPL and FP). For each collector 10 cases are randomly
selected from the experiments that were conducted in August 2008. Similarly, in
Table 4.6, verification results are given for the modeled evacuated tube solar collec-
tor (ET). The corresponding experiments were conducted in October 2008.
Table 4.4. Solar collector verification results for FPL.
G
( ) 2
W
m

T
amb
(
o
C)
( )
kg
h
m T
coll,in
(
o
C)
T
coll,out
(
o
C)
(Experiment)
T
coll,out
(
o
C)
(Model)
873.9 27.18 135.81 70.48 75.71 76.20
977.3 25.31 136.82 64.25 70.37 71.15
966.9 24.16 143.05 61.00 66.89 67.63
1000.8 27.67 135.26 75.14 81.08 81.79
997.0 26.41 143.42 64.66 70.86 71.45
1004.9 25.93 138.60 60.66 67.40 67.97
975.8 28.84 132.36 74.42 80.69 81.10
936.3 28.66 140.58 56.31 62.59 63.31
973.8 23.22 137.92 56.30 62.92 63.46
952.0 28.91 134.28 80.14 85.16 86.13
72
Table 4.5. Solar collector verification results for FP.
G
( ) 2
W
m

T
amb
(
o
C)
( )
kg
h
m T
coll,in
(
o
C)
T
coll,out
(
o
C)
(Experiment)
T
coll,out
(
o
C)
(Model)
984.9 30.20 127.59 68.93 74.31 74.73
931.6 26.09 123.78 57.78 63.41 63.70
884.8 28.59 115.81 55.20 60.88 61.41
950.9 32.36 115.70 58.38 64.44 65.20
898.0 33.69 113.74 62.97 68.44 69.24
955.5 30.54 116.39 57.18 63.63 63.97
802.9 30.70 116.42 45.01 50.40 51.18
946.3 32.83 116.58 58.19 64.23 64.96
895.5 29.97 117.34 51.26 57.04 57.79
909.5 29.13 117.04 50.56 56.48 57.21

Table 4.6. Solar collector verification results for ET.
G
( ) 2
W
m

T
amb
(
o
C)
( )
kg
h
m T
coll,in
(
o
C)
T
coll,out
(
o
C)
(Experiment)
T
coll,out
(
o
C)
(Model)
722.1 16.35 143.42 74.14 79.42 80.13
899.0 16.85 139.51 73.85 80.83 81.84
852.9 21.05 140.50 84.61 91.22 91.91
889.8 18.72 140.60 70.83 77.76 78.80
903.3 16.71 142.54 75.08 82.12 82.91
929.3 16.97 144.18 50.35 57.63 58.94
960.2 15.19 143.81 42.66 50.42 51.57
892.8 16.13 144.01 46.64 53.88 54.82
933.6 16.44 143.75 45.88 53.52 54.50
771.1 13.79 152.56 46.82 52.82 53.38

73
For Table 4.4 through Table 4.6, G is solar radiation, T
amb
is ambient temperature, m
is mass flow rate of the working fluid, T
coll,in
is inlet temperature of the collector and
T
coll,out
is outlet temperature of the collector. As can be seen from the results from
Table 4.4 through Table 4.6, all the differences of the outlet temperatures are below
1.2
o
C. Therefore, the collector model results are close to the actual results within
acceptable limits and the models are verified. In addition, the second flat plate collec-
tor (FP) gives higher thermal efficiencies and has higher stagnation temperature than
the first flat plate collector (FPL) as stated in the collectors catalog. Therefore, the
simulations are continued only using the second flat plate collector (FP).
74
CHAPTER 5

RESULTS
The information presented in this chapter is an extension of that previously published
or under revision by Taylan, Baker and Kaftanolu [44, 89, 108].
Trends in COP
ads
and COP
sys
are investigated while varying the cycle type, adsor-
bent refrigerant (working) pair and working conditions given in Table 4.1. Alt-
hough it is possible to get errors that are defined in Equations (4.5) and (4.6) on the
order of 10
-8
, the errors are kept less than 3.5% so as not to increase the simulation
times significantly. The relation between simulation times and the interval sizes of
the finite difference solution can be observed in Table 4.3 to some extent. Further-
more, convergence criterion for the finite difference scheme is selected as 10
-6
. The
convergence criterion is described as the estimated absolute error, i.e., the absolute
difference between the present result and corresponding result in the previous time
step or attempt. More than 10,000 cases were investigated, and many of the investi-
gated cases produced similar results and trends. For brevity and clarity, only the most
significant results are presented and discussed in this chapter. Additionally, the pre-
sented trends are representative of the trends for all analyzed cases.
5.1 Adsorption Cooling Cycle Comparison
Trends in COP
ads
for the thermal wave cycles with no mass recovery (NMR), adia-
batic mass recovery (AMR) and isothermal mass recovery (IMR) are given in Figure
5.1. All thermal wave cycles are also presented with or without by-pass lines for zeo-
lite NaX water (Z1) pair. Also presented in this Figure are COP
ads
values for the
reversible cycle, simple cycle and heat recovery cycle with two isothermal beds. As
required, in all cases COP
ads
of the reversible cycle is the highest since the reversible
75
cycle is the thermodynamically limiting case. From Figure 5.1, it is observed that for
the conditions explored the thermal wave cycles are superior to the simple and heat
recovery cycles, especially when the maximum bed temperature (T
hot
) is high. In
other words, as T
hot
increases, COP
ads
trends of all the thermal wave cycles and sim-
ple or heat recovery cycles diverge. Although the differences between the thermal
wave cycles with and without by-pass line among the investigated cases are very
small, COP
ads
values of the thermal wave cycles with by-pass line are slightly higher
than the corresponding values for the thermal wave cycles without by-pass line.
The working pair in Figure 5.1 is zeolite NaX water (Z1) and the corresponding
COP
ads
curves for NMR and AMR coincide while the COP
ads
curves for IMR are
slightly higher than the other thermal wave cycles COP
ads
curves. This result holds
for both thermal wave cycles with and without by-pass lines.

Figure 5.1. Comparison of different adsorption cycle types for Z1 pair, T
cond
= 30
o
C,
T
evap
= 10
o
C, R = 10 and AT
excess
= 0
o
C.

0.0
1.0
2.0
3.0
4.0
5.0
90 100 110 120 130 140 150 160 170 180
C
O
P
a
d
s

Maximum Bed Temperature, T
hot
(
o
C)
Reversible NMR with bypass NMR without bypass
AMR with bypass AMR without bypass IMR with bypass
IMR without bypass Heat Recovery Simple
76
In addition to Figure 5.1, in Figure 5.2 the COP
ads
trends for different adsorption
cooling cycles are given for an activated carbon ammonia (CA) working pair. Un-
like the results for zeolite NaX water (Z1) working pair given in Figure 5.1, COP
ads

curves of AMR and NMR do not coincide. For maximum bed temperature lower
than approximately 110
o
C (T
hot
< ~110
o
C), COP
ads
values of AMR are slightly high-
er than COP
ads
values of NMR. For maximum bed temperatures higher than approx-
imately 110
o
C (T
hot
> ~110
o
C), COP
ads
values of NMR are higher than COP
ads
values
of the AMR as shown in Figure 5.2. Considering all investigated cases, IMR gives
higher COP
ads
values than NMR and AMR.

Figure 5.2. Comparison of different adsorption cycle types for CA pair, T
cond
= 30
o
C,
T
evap
= 10
o
C, R = 10 and AT
excess
= 0
o
C.
The relative difference of COP
ads
between NMR and AMR for zeolite NaX water
varies as -0.01 < (COP
ads,AMR
COP
ads,NMR
) / COP
ads,NMR
< 0.17 and is large only
when T
hot
is low. The relative difference of COP
ads
between NMR and IMR for zeo-
lite NaX water varies as -0.01 < (COP
ads,IMR
COP
ads,NMR
) / COP
ads,NMR
< 1.62 and
is large when the condensation temperature (T
cond
) is high and T
hot
is low. Converse-
0.0
1.0
2.0
3.0
4.0
5.0
80 90 100 110 120 130 140 150
C
O
P
a
d
s

hot
(
o
C)
Reversible NMR with bypass NMR without bypass
AMR with bypass AMR without bypass IMR with bypass
IMR without bypass Heat Recovery Simple
77
ly, the relative differences of COP
ads
between all investigated thermal wave cycles
are low when T
cond
is low and T
hot
is high.
5.2 Adsorbent Refrigerant Pair Comparison
Clapeyron diagrams (ln(P
v
) vs. -1/T) for different adsorbent refrigerant (working)
pairs are given in Figure 5.3 and Figure 5.4. As it can be observed from both figures,
the activated carbon ammonia (CA) pair works with the highest vapor pressure
among the investigated working pairs given in Table 4.1, whereas activated carbon
methanol (CM) pair has the second highest vapor pressure. As is obvious from Sec-
tion 3.1, vapor pressures are related with the saturation pressures of the correspond-
ing refrigerants and the condensation and evaporation pressures. Maximum and min-
imum bed temperatures are independent of the working pairs, and they are fixed in
each simulation by user/designer as shown in Figure 5.3 and Figure 5.4. For a certain
case, bed temperatures at the end of Process 12 (state 2, end of the isosteric heat-
ing) and Process 34 (state 4, isobaric heating and desorption) differ slightly for
each working pair. The pairs with water as refrigerant (SG, Z1 and ZW) can be listed
with decreasing temperature at state 2 (T
2
) as silica gel water (SG), zeolite NaX
water (Z1) and zeolite X13 water (ZW) and with decreasing temperature at state 4
(T
4
) as zeolite X13 water (ZW), silica gel water (SG) and zeolite NaX water
(Z1). Similarly, the pairs with activated carbon as adsorbent can be listed with de-
creasing temperature at state 2 (T
2
) as activated carbon ammonia (CA) and activat-
ed carbon methanol (CM) and with decreasing temperature at state 4 (T
4
) as acti-
vated carbon methanol (CM) and activated carbon ammonia (CA).
In order to illustrate the sorption regimes described in Section 2.2.1.b, Figure 5.3 and
Figure 5.4 can be compared. Note that in Figure 5.3, the temperature at state 4 (T
4
) is
higher than temperature at state 2 (T
2
), i.e., T
4
> T
2
. Therefore, the condition given in
Figure 5.3 is suitable for paired sorption heat recovery. On the contrary, the tempera-
ture at state 4 (T
4
) is lower than the temperature at state 2 (T
2
), i.e., T
4
< T
2
in Figure
5.4. For the conditions given in Figure 5.4 the no sorption regime can occur, and as
78
explained in Section 2.2.1.b the illustrated condition in Figure 5.4 limits heat recov-
ery.

hot
= 90
o
C, T
cond
=
30
o
C, T
evap
= 10
o
C, R = 10 and AT
excess
= 0
o
C. Numbers on plot area indicate the state
numbers.

hot
= 90
o
C, T
cond
=
35
o
C, T
evap
= 5
o
C, R = 10 and AT
excess
= 0
o
C. Numbers on plot area indicate the states.
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
-0.0034 -0.0033 -0.0032 -0.0031 -0.0030 -0.0029 -0.0028 -0.0027
l
n
(
P
v
)

(
l
n

k
P
a
)

-1/T (-1/K)
Z1 ZW SG CA CM
1
2
3
4
1
2
4 3
1
2
3
4
-1.0
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
-0.0033 -0.0032 -0.0031 -0.0030 -0.0029 -0.0028 -0.0027
l
n
(
P
v
)

(
l
n

k
P
a
)

-1/T (-1/K)
Z1 ZW SG CA CM
1
1
1
2
2
2
3
3
3
4
4
4
79
Working pairs are compared in Figure 5.5 in terms of COP
ads
. For thermal wave cy-
cles, the trend in COP
ads
for all working pairs is always increasing with increasing
T
hot
as shown in Figure 5.5. For heat recovery and simple cycles, the COP
ads
curves
increase for all working pairs and COP
ads
asymptotically reach different values for
different pairs as shown in Figure 5.5. The working pair with the highest COP
ads
var-
ies with T
hot
. For the thermal wave cycles, silica gel water (SG) pair has the highest
COP
ads
for the T
hot
range of 80-100
o
C, activated carbon ammonia (CA) pair for
100-150
o
C and zeolite NaX water (Z1) pair for 150-190
o
C. For the heat recovery
and simple cycles, silica gel water (SG) pair still has the highest COP
sys,clg
for the
80-100
o
C range and zeolite X13 water (ZW) pair has the highest COP
sys,clg
for the
100-160
o
C region.

Figure 5.5. Working pair comparison using different adsorption cycle types with
T
cond
= 30
o
C, T
evap
= 10
o
C, R = 10 and AT
excess
= 0
o
C.
80
In addition to Figure 5.5, the working pairs are also compared using seasonal-
transient simulations. In Figure 5.6, the working pairs are compared in terms of
COP
sys,clg
. Similar to the results in Figure 5.5, the COP
sys,clg
trends of thermal wave
cycles are to increase with increasing T
hot
as shown in Figure 5.6. For the heat recov-
ery and simple cycles, COP
sys,clg
values increase with increasing T
hot
, achieve a peak
and then decrease. For these cycles, working pairs have peak COP
sys,clg
values at dif-
ferent T
hot
values. The peak COP
sys,clg
is at approximately 90
o
C for silica gel water
(SG) and activated carbon methanol (CM) pairs, 100
o
C for activated carbon am-
monia (CA) and zeolite X13 water (ZW) pairs, and 120
o
C for zeolite NaX water
(Z1) pair. As with the COP
ads
results, COP
sys,clg
results also show that each pair
achieves the highest COP
sys,clg
at a different T
hot
as can be seen in Figure 5.6.

Figure 5.6. Working pair comparison using different adsorption cycle types with dry
cooling tower, T
evap
= 10
o
C, R = 10 and AT
excess
= 0
o
C.
81
5.3 Design and Working Conditions Comparison
In this section, effects of the parameters that are given in Table 4.1 on performances
of the adsorption cooling cycle (COP
ads
) and the solar thermal integrated adsorption
cooling system (in terms of COP
sys
for steady simulations and COP
sys,clg
and r
sys,clg

for seasonal-transient simulations) are given in detail.
5.3.1 Effects of Collector Choice and Solar Radiation Level
As mentioned in Section 3.2.1, two types of solar thermal collectors are modeled: flat
plate and evacuated tube collectors. Although the collector choice does not affect
COP
ads
, it affects COP
sys
(and COP
sys,clg
) since COP
sys
is calculated using Equation
(4.2). Meanwhile, solar radiation (G) level has a direct effect on the efficiencies of
solar thermal collectors as can be seen from Equation (3.12).
In Figure 5.7, collector efficiencies and the simple adsorption cycles COP
ads
are
given for steady simulations as G varies. As expected, the evacuated tube collector
has higher thermal efficiency than the flat plate collector for all investigated cases.
As can be seen from Figure 5.7, thermal efficiencies of the solar thermal collectors
(q
coll
) decrease with increasing T
hot
, and under certain conditions (e.g., G =
500W/m
2
, T
cond
= 20
o
C or 35
o
C and T
hot
> 110
o
C for flat plate collector in Figure
5.7), the solar thermal collectors reach their stagnation temperature, defined as the
maximum achievable collector temperature with a stagnated fluid [112, 113], and
consequently collector efficiency becomes zero (q
coll
= 0). As shown in Figure 5.7
(and Figure 5.8), the flat plate collector reaches zero thermal efficiency at lower tem-
peratures than the evacuated tube collector, i.e., the flat plate collector has lower
stagnation temperatures than the evacuated tube collector. Therefore, the modeled
flat plate collector is not suitable for adsorption cooling cycles that require high
(driving) temperatures. Additionally, as G level increases, q
coll
and stagnation tem-
peratures for both collectors increase.
82

Figure 5.7. Collector and solar radiation level comparison for simple cycle, Z1 pair,
T
cond
= 20
o
C, T
amb
= 35
o
C, T
evap
= 10
o
C, R = 10 and AT
excess
= 10
o
C (Legend:
CC,GGG where CC = collector type: FP = flat plate; ET = evacuated tube; GGG=G
(W/m
2
)).
COP
sys
changes with variations in T
hot
, G and collector type are given in Figure 5.8.
For the investigated T
hot
range, systems with an evacuated tube collector (ET) have a
peak COP
sys
at different T
hot
values depending on the working pair as mentioned in
Section 5.2. As in Figure 5.7, stagnation temperature for flat plate collector and in-
creasing stagnation temperature with increasing G can be observed in Figure 5.8.
Moreover, COP
sys
increases with increasing G.
0.00
0.05
0.10
0.15
0.20
0.25
0%
10%
20%
30%
40%
50%
60%
70%
80%
90 100 110 120 130 140 150
C
O
P
a
d
s

q
c
o
l
l

hot
(
o
C)
FP,500 FP,750 FP,1000 ET,500
ET,750 ET,1000 COPads
83

Figure 5.8. Collector and solar radiation level comparison for simple cycle, Z1 pair,
T
cond
= 20
o
C, T
amb
= 35
o
C, T
evap
= 10
o
C, R = 10 and AT
excess
= 10
o
C (Legend:
CC,GGG where CC = collector type: FP = flat plate; ET = evacuated tube; GGG=G
(W/m
2
)).
5.3.2 Effects of Condensation Temperature
Effects of T
cond
on COP
ads
are investigated through steady simulations, and repre-
sentative results can be seen in Figure 5.9 for NMR with different working pairs. For
all working pairs and working conditions, as T
cond
increases, COP
ads
decreases. Addi-
tionally, as T
hot
increases, the differences in COP
ads
due to changes in T
cond
increase
according to steady results as can be seen in Figure 5.9. For investigated working
pairs and conditions the level of COP
ads
increase when condensation temperature
decreases from T
cond
= 35
o
C to 30
o
C is smaller than the level of COP
ads
decrease
when condensation temperature decreases from T
cond
= 30
o
C to 25
o
C, although the
decrease in condensation temperature for each case is the same and 5
o
C.
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
90 100 110 120 130 140 150
C
O
P
s
y
s

hot
(
o
C)
FP,500 FP,750 FP,1000
ET,500 ET,750 ET,1000
84

Figure 5.9. Condensation temperature comparison for different working pairs, NMR,
T
evap
= 10
o
C, R = 10 and AT
excess
= 10
o
C (Legend shows working pair and T
cond
(
o
C)).
The effect of using ideal dry (T
cond
= T
o
) versus ideal wet (T
cond
= T
wb
) cooling towers
are analyzed through seasonal-transient simulations, and representative results are
shown in Figure 5.10. In all investigated cases, systems with wet cooling towers have
higher efficiencies than systems with dry cooling towers. Unlike the steady simula-
tion results, the differences in COP
sys,clg
with varying T
cond
diminish as T
hot
increases
for simple and heat recovery cycles. The difference in COP
sys,clg
curve still increases
for thermal wave cycles. As mentioned in Sections 5.1 and 5.3.1, the systems with an
evacuated tube collector and thermal wave cycle have the highest COP
sys,clg
values
followed by systems with an evacuated tube collector and heat recovery cycle. Sys-
tems with a flat plate collector and simple cycle have the lowest COP
sys,clg
values.
0.0
1.0
2.0
3.0
4.0
5.0
80 90 100 110 120 130 140 150 160
C
O
P
a
d
s

hot
(
o
C)
ZW,25 ZW,30 ZW,35
CA,25 CA,30 CA,35
SG,25 SG,30 SG,35
85

Figure 5.10. Cooling tower comparison for CA and Z1 pairs, simple and heat recov-
ery cycles, T
evap
= 10
o
C, R = 10 and AT
excess
= 10
o
C (Legend shows the working pair,
cycle type and cooling tower type).
In addition to the COP
sys,clg
results that are obtained through seasonal-transient simu-
lations, effects of cooling tower type on COP
sys,clg
ratios (r
sys
) are also investigated
and results are presented in Figure 5.11. As for the results with COP
sys,clg
, using a wet
cooling tower increases r
sys
for all working pairs, cycle types and operating condi-
tions. As T
hot
increases, r
sys
continuously increases for the thermal wave cycles while
r
sys
asymptotically approaches to approximately 1 for the heat recovery cycle. Addi-
tionally, r
sys
approach to zero for the simple cycle since r
sys
definition is based on
simple cycle of the base case and dry cooling tower is selected for the base case.
Hence, COP
sys,clg
of simple cycles approaches to their corresponding COP
sys,clg
of the
base cases as T
hot
increases as can be derived using Equation (4.4).
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0.16
80 90 100 110 120 130 140 150 160 170 180 190
C
O
P
s
y
s
,
c
l
g

hot
(
o
C)
CA,Simple,Dry CA,HRec,Dry
CA,Simple,Wet CA,HRec,Wet
Z1,Simple,Dry Z1,HRec,Dry
Z1,Simple,Wet Z1,HRec,Wet
86

Figure 5.11. Cooling tower comparison for different adsorption cycle types, working
pairs, T
evap
= 10
o
C, R = 10 and AT
excess
= 0
o
C (Legend shows working pair and con-
denser type).
5.3.3 Effects of Evaporation Temperature
Using steady simulations, effects of evaporation temperature (T
evap
) on the perfor-
mance of the adsorption cooling system (COP
ads
) are analyzed, and representative
results are given in Figure 5.12. As can be observed from Figure 5.12, as T
evap
in-
creases, COP
ads
also increases for all the investigated cases. As with the effects of
T
cond
on COP
ads
that are mentioned in Section 5.3.2, the differences between COP
ads

curves of a particular case (i.e., adsorption cycle type, working pair, T
cond
, AT
excess

and R) due to changes in T
evap
increase with increasing T
hot
. Moreover, the increase in
COP
ads
as evaporation temperature increases from T
evap
= 10
o
C to 15
o
C is larger than
87
the increase in COP
ads
as evaporation temperature increases from T
evap
= 5
o
C to 10
o
C
for the investigated cases as can be seen in Figure 5.12.

Figure 5.12. Evaporation temperature comparison for NMR, ZW, CA and SG pairs,
T
cond
= 30
o
C, R = 10 and AT
excess
= 10
o
C (Legend shows working pair and T
evap
(
o
C)).
Using seasonal-transient simulations, effects of T
evap
on COP
sys,clg
are analyzed and
shown in Figure 5.13 for a single set of working conditions, two different adsorption
cooling cycle types and two different working pairs. As with the steady simulations,
increasing T
evap
increases the COP
sys,clg
values for the investigated cases given in
Table 4.1. The differences between COP
sys,clg
curves become smaller for simple and
heat recovery cycles as T
hot
increases as can be seen in Figure 5.13. The differences
between COP
sys,clg
curves as T
evap
changes vary with the other operating conditions,
adsorption cooling cycle type and working pairs.
0.0
1.0
2.0
3.0
4.0
5.0
80 90 100 110 120 130 140 150 160
C
O
P
a
d
s

hot
(
o
C)
ZW,5 ZW,10 ZW,15
CA,5 CA,10 CA,15
SG,5 SG,10 SG,15
88

Figure 5.13. Evaporation temperature comparison for simple and heat recovery cy-
cles, CA and Z1 pairs, T
cond
= 30
o
C, R = 10 and AT
excess
= 10
o
C (Legend shows the
working pair, cycle type and T
evap
(
o
C)).
General trends of r
sys
as T
evap
increases are almost same as changing from dry cooling
tower to wet cooling tower. r
sys
results from seasonal-transient simulations are given
in Figure 5.14. With increasing T
hot
and for all analyzed T
evap
and working pairs, r
sys

increase for the thermal wave cycles as shown in Figure 5.14 and r
sys
values ap-
proach to zero for the simple cycle as in the T
cond
cases. However, the trends for the
heat recovery cycle are slightly different. r
sys
for the heat recovery cycle with T
evap
=
15
o
C reach a minimum value and then increase as T
hot
increases, except for activated
carbon methanol (CM) pair. r
sys
values for the heat recovery cycle with activated
carbon methanol (CM) pair and T
evap
= 15
o
C increase as T
hot
increases unlike other
working pairs. Therefore, the general tendency of r
sys
is increasing for the heat re-
covery cycle, but still the differences between r
sys
values with different T
evap
values
decrease as T
hot
increases. Ratios less than zero for the simple cycle show that those
cases have COP
sys
values less than COP
sys
of the base case as can be inferred from
Equation (4.4).
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
80 90 100 110 120 130 140 150 160 170 180 190
C
O
P
s
y
s
,
c
l
g

hot
(
o
C)
CA,Simple,5 CA,Simple,10 CA,Simple,15
CA,HRec,5 CA,HRec,10 CA,HRec,15
Z1,Simple,5 Z1,Simple,10 Z1,Simple,15
Z1,HRec,5 Z1,HRec,10 Z1,HRec,15
89

Figure 5.14. Evaporation temperature comparison for different adsorption cycle types
and working pairs, T
cond
= 30
o
C, R = 10 and AT
excess
= 0
o
C (Legend shows working
pair and T
evap
(
o
C)).
5.3.4 Effects of Excess Bed Temperature
Effects of excess bed temperature (AT
excess
) on adsorption cooling cycles are ana-
lyzed through a series of steady simulations, and representative results are given in
Figure 5.15. As mentioned in Section 4.1, AT
excess
is defined as the temperature dif-
ference between the minimum bed temperature (T
o
) and the condensation tempera-
ture (T
cond
). For all the investigated cases given in Table 4.1, increasing AT
excess
shifts
the COP
ads
curve downwards on COP
ads
vs. T
hot
graph as shown in Figure 5.15. In
other words, decreasing T
o
to T
cond
is advantageous in terms of COP
ads
. Unlike the
results for T
cond
and T
evap
in Sections 5.3.2 and 5.3.3, the differences between COP
ads

90
curves with only variations in AT
excess
are almost constant. The differences between
COP
ads
values are also almost constant as T
hot
varies.

Figure 5.15. Excess bed temperature comparison for NMR and ZW, CA and SG
pairs, T
cond
= 30
o
C, T
evap
= 10
o
C and R = 10 (Legend shows working pair and AT
excess

(
o
C)).
For the steady simulations and simple or heat recovery cycles where T
o
= T
cond
= T
amb

(therefore AT
excess
= 0
o
C), at lower T
hot
values, low T
cond
values give higher COP
sys

while at higher T
hot
values, high T
cond
values give higher COP
sys
. In another set of
cases, a dry cooling tower is used (i.e., T
cond
= T
amb
) and excess bed temperature is
either AT
excess
= 10
o
C or 15
o
C. In these simulations, COP
sys
increases with increasing
AT
excess
at low T
hot
values (mostly in the region of T
hot
< 120
o
C), but decreases with
increasing T
cond
as T
hot
increases and at high T
hot
values (for approximately T
hot
>
120
o
C). For the final set of cases, T
cond
and T
o
are not equal to each other and ambi-
ent temperature is fixed at T
amb
= 35
o
C. Results of these cases are similar to the re-
sults of the previous set of cases.
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
80 90 100 110 120 130 140 150 160
C
O
P
a
d
s

hot
(
o
C)
ZW,0 ZW,5 ZW,10
CA,0 CA,5 CA,10
SG,0 SG,5 SG,10
91
According to the results of seasonal-transient simulations, when the adsorbent bed is
not cooled down to T
cond
(i.e., AT
excess
> 0), COP
sys,clg
increases as shown in Figure
5.16. Additionally, as AT
excess
increases, COP
sys,clg
increases except in one case. This
exception is observed when using a dry cooling tower with T
hot
= 90
o
C and R = 10
regardless of the collector type and adsorption cycle. In general, considering the in-
vestigated cases given in Table 4.1, the effect of AT
excess
on COP
sys,clg
is small when
compared to the effects of T
cond
and T
evap
.

Figure 5.16. Excess bed temperature comparison for simple and heat recovery cycles,
CA and Z1 pairs, T
cond
= 30
o
C, T
evap
= 10
o
C and R = 10 (Legend shows the working
pair, cycle type and AT
excess
(
o
C)).
Effects of AT
excess
are also analyzed through r
sys
that are obtained from seasonal-
transient simulations. r
sys
results are summarized in Figure 5.17. Increasing AT
excess

for thermal wave cycle decreases r
sys
for all working pairs. Additionally, as T
hot
in-
creases r
sys
increases and the differences between ratios of different AT
excess
values
decrease slightly. The general r
sys
trend for the simple and heat recovery cycles is
similar to the trend for the thermal wave cycles, except r
sys
values approach to some
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
80 90 100 110 120 130 140 150 160 170 180 190
C
O
P
s
y
s
,
c
l
g

hot
(
o
C)
92
specific values that depend on working pairs. r
sys
of the simple and heat recovery
cycles continuously increase and approach to those values as T
hot
increases as can be
seen in Figure 5.17. However, unlike other working pairs, r
sys
of the simple cycle
with zeolite NaX water (Z1) pair approach their maximum values and then de-
crease. As shown in Figure 5.17, r
sys
curves that correspond to different AT
excess
val-
ues coincide around T
hot
= 100-110
o
C for simple and heat recovery cycles with a zeo-
lite NaX water (Z1) pair.

Figure 5.17. Excess bed temperature comparison for different adsorption cycle types
and working pairs, T
cond
= 30
o
C, T
evap
= 10
o
C and R = 10 (Legend shows working
pair and AT
excess
(
o
C)).
93
5.3.5 Effects of Heat Capacity Ratio
Steady simulations are also performed to investigate the effects of the heat capacity
ratio (R) values on COP
ads
. Some of the results are given in Figure 5.18. As can be
partially observed from Figure 5.18, increasing R decreases COP
ads
of the simple and
heat recovery cycles but increases COP
ads
of the thermal wave cycles. For the inves-
tigated cases and similar to the effects of T
cond
and T
evap
given in Sections 5.3.2 and
5.3.3, the differences between COP
ads
curves increases as R varies and T
hot
increases.
Additionally, in the steady cases for the zeolite NaX water (Z1) pair, heat recovery
cycle and evacuated tube collector, decreasing the heat capacity ratio from R = 10 to
R = 0 while keeping all other variables constant causes COP
sys
values to increase 2.7
6.5 times. Similar trends are obtained using flat plate collector, and these results are
consistent with previous studies [90] as are the results for variations in COP
sys
with
variations in T
hot
.

Figure 5.18. Heat capacity ratio comparison for NMR, ZW, CA and SG pairs, T
cond
=
30
o
C, T
evap
= 10
o
C and AT
excess
= 0
o
C (Legend shows working pair and R).
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
80 90 100 110 120 130 140 150 160
C
O
P
a
d
s

hot
(
o
C)
ZW,0 ZW,3 ZW,10
CA,0 CA,3 CA,10
SG,0 SG,3 SG,10
94
Moreover, representative results of seasonal-transient simulation are given in Figure
5.19. From Figure 5.19, as R increases for the simple cycle, COP
sys,clg
decreases sig-
nificantly in all cases, and adding heat recovery increases the COP
sys,clg
values
around two times when compared with the simple cycle. However, increasing R in-
creases COP
sys,clg
values for thermal wave cycles except for a few cases. These ex-
ceptions are observed when thermal wave cycles with zeolite NaX water pair is
used and when the maximum bed temperatures are below approximately (T
hot
<
~150
o
C). For these exceptional cases, increasing R decreases COP
sys,clg
as with the
simple and heat recovery cycles.

Figure 5.19. Heat capacity ratio comparison for simple and heat recovery cycles, CA
and Z1 pairs, T
cond
= 30
o
C, T
evap
= 10
o
C and AT
excess
= 10
o
C (Legend shows the work-
ing pair, cycle type and R).
Although R does not have a strong effect on the COP
sys
of the thermal wave cycles,
it does have a strong effect on COP
sys,clg
ratios since R has a strong influence on the
simple and heat recovery cycles [44]. r
sys
values that correspond to different R values
are given in Figure 5.20. For the investigated T
hot
range and the thermal wave cycles,
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
80 90 100 110 120 130 140 150 160 170 180 190
C
O
P
s
y
s
,
c
l
g

hot
(
o
C)
95
COP
sys,clg
and r
sys
increase and the differences between r
sys
increase as T
hot
increases
for all working pairs, except for the zeolite NaX water (Z1) pair. For the thermal
wave cycles with zeolite NaX water (Z1) pair, as with the COP
sys,clg
results, r
sys
of
different R values coincide at approximately T
hot
= 150
o
C. Cases with zeolite NaX
water (Z1) pair and R = 0 have higher r
sys
values for T
hot
< 150
o
C while for T
hot
>
150
o
C this case has lower r
sys
values among the other cases with zeolite NaX water
(Z1) pair and various R values. The r
sys
trends of working pairs with zeolite as adsor-
bent (zeolite NaX water (Z1) and zeolite X13 water (ZW) pairs) for simple and
heat recovery cycles as R decreases are different than the other working pairs. As T
hot

increases, r
sys
curves of these pairs (Z1 and ZW) decrease and then increase after
some T
hot
values that depend on the working conditions and pair. However, the heat
recovery cycle with silica gel water (SG), activated carbon ammonia (CA) and
activated carbon methanol (CM) pairs have continuously increasing trends as T
hot

increases for all R values. For the simple cycles with various R values, the r
sys
trends
are similar to the heat recovery cycles trends of zeolite NaX water (Z1) and zeolite
X13 water (ZW) pairs. Nevertheless, as R increases, r
sys
values decrease for all
simple and heat recovery cycles.
5.3.6 Effects of Maximum Bed Temperature
From Figure 5.1 through Figure 5.20, all results are given based on maximum bed
temperature (T
hot
). As shown in these figures, T
hot
has a direct effect on COP
ads
,
COP
sys,clg
and r
sys
. As T
hot
increases for steady simulations, COP
ads
also increases for
all the investigated cases including the different adsorption cycle types and working
pair types. Seasonal-transient simulations show that COP
sys,clg
of the thermal wave
cycles are approaching to COP
sys,clg
of the reversible cycle as T
hot
increases for all
working pairs as shown in Figure 5.5. Furthermore, COP
sys,clg
of the simple and heat
recovery cycles increase until a maximum COP
sys,clg
value is reached, then decrease
with increasing T
hot
.
96

Figure 5.20. Heat capacity ratio comparison for different adsorption cycle types and
working pairs, T
cond
= 30
o
C, T
evap
= 10
o
C and AT
excess
= 0
o
C (Legend shows working
pair and R).
5.3.7 Summary of Effects of Investigated Parameters
Results as different parameters, adsorption cycle types and working pairs vary are
presented in terms of COP
sys
for steady simulations and COP
sys,clg
and r
sys
for transi-
ent simulations. Before concluding this subsection, it would be beneficial to present
and compare the effects of investigated parameters representatively.
The results for the variation of the parameters given in Table 4.1 are shown in Figure
5.21 for IMR with by-pass line, with these results being typical of the other thermal
wave cycles. As T
hot
increases, COP
ads
increases in all cases. Of investigated temper-
atures and for a fixed temperature change, T
cond
has the largest effect on COP
ads

97
while T
evap
has the second largest effect for the thermal wave cycles. COP
ads
decreas-
es with increasing T
cond
but increases with increasing T
evap
. Increasing AT
excess
shifts
the COP
ads
curve downward in Figure 5.21. The effect of R on COP
ads
is more com-
plicated than the other parameters, since the effect of R depends on the adsorption
cycle type. Increasing R decreases COP
ads
for the simple and heat recovery cycles
but increases COP
ads
for the thermal wave cycles except for the zeolite NaX water
pair with maximum bed temperature, T
hot
< 150
o
C as shown in Figure 5.21. Howev-
er, unlike for the simple and heat recovery cycles, variations in AT
excess
and more
importantly R do not affect the thermal wave cycles COP
ads
significantly.
The changes of COP
ads
for the simple and heat recovery cycles in absolute values are
not as large as the changes for the thermal wave cycles. Additionally, for the simple
and heat recovery cycles, the investigated parameters have the same effects on the
COP
ads
values as for the thermal wave cycles. However, among all the investigated
parameters, the effect of R on COP
ads
for the simple and heat recovery cycles is the
largest unlike the thermal wave cycles.

ads
for IMR and Z1 pair (Leg-
end shows deviations from the base case which is selected as T
cond
= 30
o
C, T
evap
=
10
o
C, R = 3 and AT
excess
= 5
o
C).
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
90 100 110 120 130 140 150 160 170 180
C
O
P
a
d
s

hot
(
o
C)
base Tcond=20 Tcond=40
Tevap=5 Tevap=15 DTexcess=0
DTexcess=10 Tcond=20, R=0 Tcond=20, R=10
98
In Figure 5.22, COP
sys,clg
values are given as investigated parameters and cycles vary
for zeolite X13 water (ZW) pair. The results of COP
ads
are also applicable to the
results of COP
sys,clg
. The difference between the results of COP
ads
and COP
ads,clg
is
mainly due to the efficiency of the modeled solar thermal collectors as can be in-
ferred from Equations (4.2) and (4.3). The general trend of COP
sys,clg
curves of the
thermal wave cycles is to increase while the trend for the simple and heat recovery
cycles are to increase, achieve a maximum value and then decrease with increasing
T
hot
. Additionally, the order of the parameters from the highest to lowest effect on
COP
sys,clg
is the same as in COP
ads
results.
5.4 Normalized Model Results
Seasonal-transient simulations using the normalized model are performed for the
simple and heat recovery cycles but not for the thermal wave cycles, and effects of
the normalized storage capacity (q
Storage,max
), size of the system (S) and the other pa-
rameters listed in Table 4.1 on the solar (f) and loss (l) fractions, normalized collector
area (A
coll,N
) and normalized adsorbent mass (m
ads,N
) are analyzed and results are giv-
en in this section.
The effects of q
Storage,max
on f and l are investigated for the simple and heat recovery
cycles, and representative results for the user defined size of the cooling system, S =
1 are given in Figure 5.23 for f and in Figure 5.24 for l. Note if any net changes in
storage between the start and end of a simulation are neglected, 1 f is proportional
to the required backup power. For q
Storage,max
> 10 and S = 1, as T
hot
increases l de-
creases and becomes zero at smaller q
Storage,max
, and f also decreases but not signifi-
cantly. Systems with evacuated tube collector result in higher f and l than systems
with flat plate collector except when T
hot
= 90
o
C. When simple and heat recovery
cycles are compared, f and l do not change significantly. Additionally, using a wet
cooling tower increases f when compared to a dry cooling tower. The effect of cool-
ing tower type on l is different for the flat plate and evacuated tube collectors. For
wet cooling towers l is larger when flat plate collectors are used while for dry cooling

99

sys,clg
for different adsorption
cycle types and ZW pair (Legend shows deviations from the base case which is se-
lected as dry cooling tower, T
evap
= 10
o
C, R = 10 and AT
excess
= 0
o
C).
towers l is larger when evacuated tube collectors are used except when q
Storage,max
= 0.
Reducing R increases f and reduces l. Moreover, as AT
excess
increases, f decreases for
all investigated q
Storage,max
and l increases for all investigated q
Storage,max
3. Changes
in f and l are more significant when the heat capacity ratio changes from R = 0 to R =
10 than when the excess bed temperature changes from AT
excess
= 0
o
C to AT
excess
=
10
o
C. It is worth noting that for the flat plate collector and maximum bed tempera-
ture at T
hot
= 180
o
C, f and l are zero for all q
Storage,max
since the modeled flat plate col-
100
lector cannot achieve 180
o
C (i.e., stagnation temperatures are reached) and conse-
quently q
clg,N
= 0.

Storage,max
hot
, S = 1, simple cycle,
wet cooling tower, R = 10, AT
excess
= 10
o
C (Legend: CC,HH where CC = collector
type: FP = flat plate; ET = evacuated tube; HH=T
hot
(
o
C)).
When the effects of S on f and l are compared, as S increases f increases and l de-
creases as seen representatively in Figure 5.25 for f and in Figure 5.26 for l. For S <
4, as S increases the change in f becomes less significant as q
Storage,max
increases.
Conversely, as S increases changes in l become more significant as q
Storage,max
in-
creases. Additionally, the benefit of increasing S by a fixed increment decays as S
increases. As with the S = 1 cases, f and l of simple and heat recovery cycles do not
differ significantly, and f and l decrease as the excess bed temperature increases from
AT
excess
= 0
o
C to 10
o
C. When cases with R = 0 are compared to cases with R = 10
while S is varying, f and l are lower for R = 10 cases. For those cases, the effects of R
are more significant on the simple cycle than the heat recovery cycle.
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0 5 10 15 20
S
o
l
a
r

F
r
a
c
t
i
o
n
,

f

q
Storage,max
ET,90 ET,120 ET,150 ET,180
FP,90 FP,120 FP,150
101

Storage,max
hot
, S = 1, simple cycle,
wet cooling tower, R = 10, AT
excess
= 10
o
C (Legend: CC,HH where CC = collector
type: FP = flat plate; ET = evacuated tube; HH=T
hot
(
o
C)).

Storage,max
for various S values, flat plate collector, simple cycle, T
hot
= 90
o
C, wet cooling tower, and R = 10.
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0 5 10 15 20
L
o
s
s

F
r
a
c
t
i
o
n
,

l

q
Storage,max
ET,90 ET,120 ET,150 ET,180
FP,90 FP,120 FP,150
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0 5 10 15 20
S
o
l
a
r

F
r
a
c
t
i
o
n
,

f

q
Storage,max
S=1 S=2 S=3 S=4 S=5
102

Storage,max
for various S values, flat plate collector, simple cycle, T
hot
= 90
o
C, wet cooling tower, and R = 10.
Representative results for the normalized collector area (A
coll,N
) are given in Figure
5.27. In all cases, A
coll,N
has a peak value around a maximum bed temperature of T
hot

= 150
o
C, and it does not change significantly with AT
excess
. For the range of investi-
gated conditions, the parameters listed in order from the strongest to weakest effect
on A
coll,N
are collector type, adsorption cycle type, cooling tower type and R. The
required A
coll,N
is higher in systems with flat plate collector than systems with evacu-
ated tube collector as seen in Figure 5.27. Additionally, it can be observed from Fig-
ure 5.27 that using the heat recovery cycle and/or a wet cooling tower results in
smaller A
coll,N
. Lastly, in contrast to COP
sys
(or COP
sys,clg
), A
coll,N
increases with in-
creasing R.
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0 5 10 15 20
L
o
s
s

F
r
a
c
t
i
o
n
,

l

q
Storage,max
S=1 S=2 S=3 S=4 S=5
103

Figure 5.27. A
coll,N
vs. T
hot
(
o
C) for different collector, adsorption cycle, cooling tower
types and R and AT
excess
= 0
o
C (Legend: CC,AAAA,DDD,R where CC=collector
type: FP = flat plate; ET = evacuated tube; AAAA=cycle: HRec = heat recovery;
Simple = simple; DDD=cooling tower type: Dry or Wet).
Normalized mass of adsorbent (m
ads,N
) decreases as T
hot
increases and this decrease is
nearly linear for T
hot
> 120
o
C as shown in Figure 5.28. Similar to COP
sys,clg
results,
the required m
ads,N
increases with AT
excess
, and systems with a wet cooling tower re-
quire smaller m
ads,N
than systems with a dry cooling tower. Although not shown in
Figure 5.28, m
ads,N
is independent of the adsorption cycle type, collector type or R
since m
ads,N
is only a function of the adsorption capacity change for a half cycle, AX
= X
max
X
min
= X(T
o
,T
evap
) X(T
hot
,T
cond
).
1
10
100
1000
90 120 150 180
N
o
r
m
a
l
i
z
e
d

C
o
l
l
e
c
t
o
r

A
r
e
a
,

A
c
o
l
l
,
N

hot
(
o
C)
FP,Simple,Dry,R=10 FP,Simple,Wet,R=10 FP,HRec,Dry,R=10
FP,HRec,Wet,R=10 FP,Simple,Dry,R=0 FP,HRec,Wet,R=0
ET,Simple,Dry,R=10 ET,HRec,Dry,R=10
104

Figure 5.28. m
ads,N
vs. T
hot
(
o
C) for evacuated tube collector, simple and heat recovery
cycles, different cooling towers and T
o
and R = 10 (Legend: Dry or Wet cooling tow-
er + AT
excess
(
o
C)).
0.0
0.5
1.0
1.5
2.0
2.5
90 120 150 180
N
o
r
m
a
l
i
z
e
d

A
d
s
o
r
b
e
n
t

M
a
s
s

m
a
d
s
,
N

hot
(
o
C)
Dry+10degC Dry+5degC Dry
Wet+10degC Wet+5degC Wet
105
CHAPTER 6

DISCUSSIONS
Stand alone and solar-thermal system integrated adsorption cooling cycles are ana-
lyzed as several operating conditions (T
hot
, T
cond
, T
evap
, R and AT
excess
), some design
parameters (cooling tower and solar thermal collector types) and adsorbent refrig-
erant (working) pairs (Z1, ZW, SG, CA and CM) are varied. Within this study, dif-
ferent ideal adsorption cycle types are thermodynamically modeled using MATLAB
and these cycles are coupled with previously presented adsorbent refrigerant pair
models. The analyzed adsorption cycles include simple, heat recovery with two spa-
tially isothermal beds, thermal wave cycle with no mass recovery, thermal wave cy-
cle with adiabatic mass recovery and thermal wave cycle with isothermal mass re-
covery. The investigated thermal wave cycles are also divided into two types as
thermal wave cycles with by-pass line and thermal wave cycles without by-pass line.
All these adsorption cooling cycles are compared with a reversible cycle which is
considered as the thermodynamically limiting case. Note that thermal wave adsorp-
tion cooling cycles with and without enhancements are presented. Two limiting mass
recovery cases for the thermal wave cycles are considered: adiabatic mass recovery
and isothermal mass recovery. The actual mass recovery process should be a process
between adiabatic mass recovery and isothermal mass recovery.
The effects of conditions and parameters are analyzed through a series of simula-
tions. These simulations can be categorized as steady and seasonal-transient simula-
tions. All the simulations are run using TRNSYS software. In the seasonal-transient
simulations, hourly weather data for the city of Antalya on Turkeys Mediterranean
coast is used with 15-minute intervals for the summer period from June 1 to Septem-
ber 30. Water is selected as the heat transfer fluid for all the simulations.
106
The simulation results are obtained in terms of COP
ads
and COP
sys
for steady simula-
tions and COP
sys,clg
and COP
sys,clg
ratios (r
sys
) for seasonal-transient simulations.
Among these results, r
sys
show the opportunity to improve the selected base case of
the simple adsorption cycle by introducing two spatially isothermal bed heat recov-
ery or thermal wave cycles or by altering other operating conditions. Additionally, a
normalized model is developed to investigate the modeled solar-powered adsorption
cooling system in terms of solar and loss fractions, collector area and mass of adsor-
bent as the aforementioned parameters, maximum storage capacity and size of the
system are varied. The normalized model is applied only to the simple and heat re-
covery cycles with zeolite NaX water pair.
For all analyses, for simplicity, ideal systems are assumed to investigate (thermody-
namic) limits to the energy performance, general trends and relative values. The
models are used to make relative comparisons of different adsorption cycles, quanti-
fy limits to corresponding COP
ads
, COP
sys
, COP
sys,clg
and r
sys
values, and provide
directions for more detailed research. Therefore, absolute values should not be in-
ferred from the results presented in this study.
Adsorbent refrigerant (working) pairs are modeled using the formulations present-
ed in the literature. However, these formulations do not generally include their corre-
sponding limitations, i.e., the temperature or pressure range where the formulation is
valid. Therefore, all the simulations are checked using the 2
nd
Law of Thermodynam-
ics after applying the criteria mentioned in Section 4.1 and overall energy balance
(the 1
st
Law of Thermodynamics) checks mentioned in Section 4.2. In fact, a small
number of simulation results contradict with the 2
nd
Law of Thermodynamics. These
contradictions are believed to originate from the adsorption capacity (X) models that
are selected since the overall and component-based energy checks are satisfactory for
the corresponding simulations.
In all investigated cases, the reversible cycle has the highest COP
ads
(or COP
sys
or
COP
sys,clg
) while the thermal wave cycles is second, the heat recovery is third and the
simple cycle is the last as COP
ads
(or COP
sys
or COP
sys,clg
) values are ranked from the
107
highest to the lowest. COP
sys,clg
values of the thermal wave cycles are approximately
36-93% of corresponding COP
sys
values of the reversible cycles. Considering all
working pairs, this proportion for the heat recovery cycle is around 3-42% and for
the simple cycle 1.5-33.5% that of a reversible cycle. The deviation of the thermal
wave cycles COP
ads
from the reversible case is directly related to the shift in the
beds outlet temperatures from the minimum bed temperature and maximum bed
temperature caused by the sorption processes. Conversely, the irreversibility in the
heat recovery cycle is mainly due to the heat transfer between the two beds at differ-
ent temperatures.
Adding mass recovery to a thermal wave cycle increases both the total heating load
and the total heat recovered. For the conditions considered, the increases in total
heating load and heat recovery are similar resulting in little change in the cycles
COP
ads
. The results showed that if the aim is to increase COP
ads
, implementing mass
recovery on thermal wave cycle is not essential when the complications of mass re-
covery process, like additional control instrumentation and physical connection of
beds, are considered. However, adding mass recovery may be appropriate to meet
other goals such as shortened cycle times and increased specific cooling power. For
completeness of the effects of mass recovery, implementing isothermal mass recov-
ery slightly increases COP
ads
of the thermal wave cycle relative to the cycle with no
mass recovery while the effects of implementing adiabatic mass recovery depend on
the other operating conditions. Additionally, for the modeled thermal wave cycles,
introducing a by-pass line also increases COP
ads
slightly.
When adsorbent refrigerant (working) pairs are compared, each pair has the highest
COP
sys,clg
(or COP
sys
) values for different maximum bed temperature range. If maxi-
mum bed temperature is around 80-100
o
C where all working pair models are appli-
cable, a system with silica gel water pair has the highest COP
sys,clg
followed by
activated carbon methanol pair. For maximum bed temperature between 100
o
C and
150
o
C, a system with activated carbon ammonia pair has the highest COP
sys,clg
for
the thermal wave cycles and lowest COP
sys,clg
for the simple and heat recovery cycles
in which systems with zeolite X13 water pair achieve the highest COP
sys,clg
in the
108
same region for the simple and heat recovery cycles. For maximum bed temperature,
T
hot
> 160
o
C, only the zeolite NaX water model is applicable.
For the solar-powered adsorption cooling systems, systems with evacuated tube col-
lector give higher COP
sys
(or COP
sys,clg
) than systems with flat plate collector as also
stated by alar et al. [114]. Additionally, as is obvious, an increasing solar radiation
level increases the system performance (in terms of COP
sys
or COP
sys,clg
) as collector
efficiencies increase and the amount of heat transfer that drives the adsorption cool-
ing cycle increases. Among the investigated operating conditions, for thermal wave
cycles the parameter that has the highest effect on COP
sys
(or COP
sys,clg
) is cooling
tower type (or condensation temperature) while evaporation temperature has the se-
cond highest effect and the heat capacity ratio and excess bed temperature has little
effect. For the simple and heat recovery cycles, the parameters that have the largest
effect on COP
sys
(or COP
sys,clg
) are, in decreasing order, the heat capacity ratio, con-
densation temperature and evaporation temperature. These results are valid for all
working pairs. In general, using a wet cooling tower instead of a dry cooling tower,
increasing evaporation temperature and decreasing excess bed temperature increases
COP
ads
of all cycle types for all working pairs. In addition, increasing the heat capac-
ity ratio decreases COP
ads
of the simple and heat recovery cycles while it increases
COP
ads
of thermal wave cycle slightly regardless of working pair. All these effects of
the parameters on COP
ads
are consistent with previous studies [56, 63, 71, 73, 75].
However, the effect of maximum bed temperature on COP
sys,clg
contradicts with the
presented results by Wang et al. [55] who only investigated a small portion of the
maximum bed temperature range presented in this study. As indicated in Section
5.3.6, COP
sys,clg
increases, achieves a peak and then decreases as maximum bed tem-
perature increases which is also consistent with results of Alam et al. [80].
According to the results of normalized model simulations, backup power is always
necessary (i.e., f < 1) to meet the cooling demand in Antalya. On the other hand, the
amount of the backup power can be reduced significantly by using an adsorption
cycle with thermal regeneration (heat recovery adsorption cycle), evacuated tube
collector and a larger storage unit (i.e., higher q
Storage,max
).
109
Moreover, zeolite NaX water pair requires a relatively high maximum bed tem-
perature that cannot be reached easily using flat plate collectors due to their lower
stagnation points. This result indicates the necessity of using evacuated tube collec-
tors for the analyzed solar-thermal system integrated adsorption cooling system. To
have the smallest collector area among the investigated cases, the system should be
operated at a low maximum bed temperature (approximately 90
o
C) which, in return,
requires large adsorbent mass. In contrast, to have minimum adsorbent mass, the
adsorption capacity swing of the half cycle should be high which corresponds to high
maximum bed temperatures values since condensation temperature (or minimum bed
temperature) is fixed by the weather data. Additionally, to have smaller backup pow-
er (and higher solar fraction), normalized collector area or mass of adsorbent (or to
have higher COP
sys,clg
), wet cooling towers are preferred over dry cooling towers if
water consumption is not a problem.
110
CHAPTER 7

CONCLUSIONS
In this study, performance trends of a solar-thermal powered adsorption cooling sys-
tem are presented using the thermodynamic models of ideal adsorption cooling cy-
cles and explicitly modeled adsorbent refrigerant (working) pairs and solar thermal
system. These three models are integrated to form the investigated solar-thermal
powered adsorption cooling system.
According to the obtained results of analyzed conditions, to have high performance
in terms of coefficient of performance, the following configuration should be pre-
ferred,
- Evacuated tube collector should be chosen rather than flat plate collector
since evacuated tube collector has higher thermal efficiency and stagnation
temperature. Therefore, high maximum bed temperatures can be achieved.
- Wet cooling tower should be chosen rather than dry cooling tower as decreas-
ing condensation temperature increases coefficient of performance of the cy-
cle.
- High evaporation temperature should be preferred although it depends on the
application area and cooling requirements.
- Adsorbent bed should be cooled down to condensation temperature as much
as possible, i.e., excess bed temperature should be minimized.
- The simple and heat recovery cycles should be designed with a small dead
mass (i.e., small heat capacity ratio) while dead mass is not so important for
the thermal wave cycles.
- The modeled solar-thermal powered adsorption cooling system is simulated
for Antalya, a city on the Mediterranean coast of Turkey where summers are
111
long and hot and backup power should be considered to meet the cooling de-
mand unless very large collector areas are used.
- Adsorbent refrigerant (working) pair choice depends strongly on the tem-
perature of the available energy source and the designed/desired maximum
bed temperature since each pair dominates different maximum bed tempera-
ture ranges.
- Although adding heat and mass recovery to the adsorption cooling cycle in-
creases the coefficient of performance of the cycle, it also increases the com-
plexity of the design of the cycle. Therefore, the choice of implementing heat
and mass recovery to the adsorption cycles should be carefully chosen ac-
cording to the needs and the desired operating conditions.
As a final remark, it is believed that this study forms a basis and shows the basic de-
velopment guidelines for a very promising technology which is the solar-thermal
powered adsorption cooling systems. In the future with appropriate and required de-
velopments and research, these systems will be commercially and commonly used
and the cooling and refrigeration will almost be free of externally supplied electrici-
ty.
112
CHAPTER 8

SUGGESTIONS FOR FUTURE
INVESTIGATIONS
This presented study should be considered a starting point for further studies. Alt-
hough the current models present the general trends of the system performance in
terms of coefficient of performance as the investigated parameters are varied, a num-
ber of improvements can be accomplished to improve the models and investigate the
performance of the system in more detail. These improvements may include the fol-
lowing items,
- The current adsorption cycle models only include thermodynamic relations.
These models can be improved by implementing heat and mass transfer and
diffusion equations based on the specific thermal design of the adsorbent bed.
As a result, some new important aspects, such as cycle time and specific cool-
ing power, can be introduced.
- In order to enhance the current model and investigate the adsorption cycle in
detail, two different approaches are suggested. One of them is to enhance the
current thermodynamic model of the entire adsorption cooling cycle to in-
clude heat and mass transfer equations in MATLAB. The other approach is to
use built-in TRNSYS components and model only the adsorption cycle in-
cluding heat and mass transfer equations and consequently, to obtain a com-
plete adsorption cooling cycle model in TRNSYS. The former approach may
be more effective for steady analysis while the latter one is believed to be
easier to implement for transient analysis. These further analyses can be done
based on components rather than system. Therefore the effect of each com-
ponent and their corresponding parameters on the system performance can be
further distinguished.
113
- The current models which include only energy relations (the 1
st
Law of
Thermodynamics) can be extended with the exergy (the 2
nd
Law of Thermo-
dynamics) analysis to investigate the improvement areas on both system- and
component-wise.
- The numerical models should be supported with experimental studies. Proto-
types of the modeled cycles should be constructed and analyzed to further
verify the results of numerical models. The emphasis in the design should be
given to the design of adsorbent bed to assure the adsorption and heat transfer
characteristics are as close as to the limiting cases provided in the present
study.
- In the present study, the thermal wave cycles seem to be a promising technol-
ogy to meet the cooling demands without introducing significant electrical
loads, but in the present study only a square wave is considered for all ther-
mal wave cycles which is a critical assumption. Therefore, some other kinds
of waves (e.g., ramp wave) can be modeled to see the performance change
according to the ideal square wave assumption.
- The results also suggest that in the thermal wave cycles, adsorbent beds with
large dead mass can be used. For this reason, engineering effort should be put
into designing the bed to approximate a square-thermal wave (which is the
limiting case) or to minimize the width of the thermal wave, possibly at the
expense of trying to minimize the beds dead mass.
- New adsorption cycle models with heat and/or mass recovery can be mod-
eled. The number of beds in these models can be varied from two to infinity
and the effect of heat and mass recovery can be analyzed as the number of
adsorbent beds is varied by defining new performance variables such as heat
and/or mass recovery efficiencies.
- Further studies can also be directed to the availability of different adsorbent
refrigerant pairs to heat and mass recovery enhancements using steady simu-
lations.
114
- New adsorbent refrigerant pairs that are available in the literature can be
added to the current models. These pairs may include some hybrid adsorbents
such as SiO
2
/ CaCl
2
.
115
BIBLIOGRAPHY

[1] Conti, J. J., Sweetnam, G. E., and Doman, L. E., "International Energy Outlook
2009," Energy Information Administration, U.S. Department of Energy,
Washington, 2009.
[2] Dhar, P. L., and Singh, S. K., "Studies on Solid Desiccant Based Hybrid Air-
Conditioning Systems," Applied Thermal Engineering, vol. 21, no. 2, pp. 119-
134, 2001.
[3] "Ongoing Research Relevant for Solar Assisted Air Conditioning Systems,"
International Energy Agency, Solar Heating and Cooling Programme, Freiburg,
2002.
[4] Calm, J. M., "Emissions and Environmental Impacts from Air-Conditioning
and Refrigeration Systems," International Journal of Refrigeration, vol. 25, no.
3, pp. 293-305, 2002.
[5] McMullan, J. T., "Refrigeration and the Environment - Issues and Strategies
for the Future," International Journal of Refrigeration, vol. 25, no. 1, pp. 89-
99, 2002.
[6] Papadopoulos, A. M., Oxizidis, S., and Kyriakis, N., "Perspectives of Solar
Cooling in View of the Developments in the Air-Conditioning Sector,"
Renewable and Sustainable Energy Reviews, vol. 7, no. 5, pp. 419-438, 2003.
[7] "Annual Development of Turkeys Gross Electricity Generation-Imports-
Exports and Demand (1975-2008)," Turkish Electricity Transmission Co.
(TEA), Ankara, 2008.
[8] "The Distribution of Gross Electricity Generation by Primary Energy
Resources and the Electric Utilities in Turkey," Turkish Electricity
Transmission Co. (TEA), Ankara, 2007.
[9] "Turkish Electrical Energy 10-Year Generation Capacity Projection," Turkish
Electricity Transmission Co. (TEA), Ankara, 2009.
[10] Naukkarinen, P., "Solar Air Conditioning and Its Role in Alleviating the
Energy Crisis of the Mediterranean Hotels," International Journal of
Sustainable Energy, vol. 28, no. 1, pp. 93 - 100, 2009.
[11] Ba, K., Antalya'da Yeni Enerji Yatrmlarna Ihtiya Duyuluyor, Last accessed
March 13, 2010, http://www.zaman.com.tr/haber.do?haberno=573763 (in
Turkish).
116
[12] Lewis, N. S., Crabtree, G., Nozik, A. J., Wasielewski, M. R., and Alivisatos, P.,
"Basic Research Needs for Solar Energy Utilization," U.S. Department of
Energy, 2005.
[13] Weiss, W., Bergmann, I., and Faninger, G., "Solar Heat Worldwide: Markets
and Contribution to the Energy Supply 2006," International Energy Agency -
Solar Heating & Cooling Programme, Gleisdorf, 2008.
[14] Weiss, W., Bergmann, I., and Faninger, G., "Solar Heat Worldwide: Markets
and Contribution to the Energy Supply 2007," International Energy Agency -
Solar Heating & Cooling Programme, Gleisdorf, 2009.
[15] Abu Hamdeh, N. H., and Al-Muhtaseb, M. T. A., "Optimization of Solar
Adsorption Refrigeration System Using Experimental and Statistical
Techniques," Energy Conversion and Management: In Press, Corrected Proof,
2010.
[16] Bilgi, M., Grep, P., Trker, L., Yntem, E., Aydnalp Kksal, M., Baker, D.
K., and Erdil, E., "Modeling of Hourly Short and Long Term Electricity
Demand at Load Distribution Regions," The Scientific and Technological
Council of Turkey (TBTAK) Project No: 108M001, Ankara, 2009.
[17] Sumathy, K., Yeung, K. H., and Yong, L., "Technology Development in the
Solar Adsorption Refrigeration Systems," Progress in Energy and Combustion
Science, vol. 29, no. 4, pp. 301-327, 2003.
[18] Wang, R., and Wang, L., "Adsorption Refrigeration-Green Cooling Driven by
Low Grade Thermal Energy," Chinese Science Bulletin, vol. 50, no. 3, pp. 193-
204, 2005.
[19] Ziegler, F., "State of the Art in Sorption Heat Pumping and Cooling
Technologies," International Journal of Refrigeration, vol. 25, no. 4, pp. 450-
459, 2002.
[20] Moran, M. J., and Shapiro, H. N., Fundamentals of Engineering
Thermodynamics, Fifth Ed., Wiley, 2006.
[21] Sonntag, R. E., Borgnakke, C., and Van Wylen, G. J., Fundamentals of
Thermodynamics, Sixth Ed., John Wiley & Sons, Inc., 2003.
[22] Sar, O. G., "Exergy Analysis of a Solar Assisted Absorption Heat Pump for
Floor Heating System," Master of Science, Department of Mechanical
Engineering, Middle East Technical University, 2004.
[23] Trott, A. R., and Welch, T., Refrigeration and Air-Conditioning, Third Ed.,
Butterworth Heinemann, Boston, 2000.
117
[24] Wang, L. W., Wang, R. Z., and Oliveira, R. G., "A Review on Adsorption
Working Pairs for Refrigeration," Renewable and Sustainable Energy Reviews,
vol. 13, no. 3, pp. 518-534, 2009.
[25] Meunier, F., "Adsorptive Cooling: A Clean Technology," Clean Technologies
and Environmental Policy, vol. 3, no. 1, pp. 8-20, 2001.
[26] Alefeld, G., and Radermacher, R., Heat Conversion Systems, CRC Press, Boca
Raton, 1994.
[27] Herold, K. E., Radermacher, R., and Klein, S. A., Absorption Chillers and Heat
Pumps, CRC Press, Boca Raton, 1996.
[28] Waugaman, D. G., Kini, A., and Kettleborough, C. F., "A Review of Desiccant
Cooling Systems," Journal of Energy Resources Technology, vol. 115, no. 1,
pp. 1-8, 1993.
[29] Henning, H. M., Erpenbeck, T., Hindenburg, C., and Santamaria, I. S., "The
Potential of Solar Energy Use in Desiccant Cooling Cycles," International
Journal of Refrigeration, vol. 24, no. 3, pp. 220-229, 2001.
[30] Daou, K., Wang, R. Z., and Xia, Z. Z., "Desiccant Cooling Air Conditioning: A
Review," Renewable and Sustainable Energy Reviews, vol. 10, no. 2, pp. 55-
77, 2006.
[31] ASHRAE, Desiccant Cooling and Dehumidification, American Society of
Heating, Refrigerating and Air-Conditioning Engineers, Atlanta, 1992.
[32] Demir, H., Mobedi, M., and lk, S., "A Review on Adsorption Heat Pump:
Problems and Solutions," Renewable and Sustainable Energy Reviews, vol. 12,
no. 9, pp. 2381-2403, 2008.
[33] Dieng, A. O., and Wang, R. Z., "Literature Review on Solar Adsorption
Technologies for Ice-Making and Air-Conditioning Purposes and Recent
Developments in Solar Technology," Renewable and Sustainable Energy
Reviews, vol. 5, no. 4, pp. 313-342, 2001.
[34] Lambert, M. A., and Jones, B. J., "Review of Regenerative Adsorption Heat
Pumps," Journal of Thermophysics and Heat Transfer, vol. 19, no. 4, pp. 471-
485, 2005.
[35] Baker, D. K., and Kaftanolu, B., "Limits to the Thermodynamic Performance
of a Thermal Wave Adsorption Cooling Cycle," Proceedings of International
Conference on Heat Transfer, Fluid Mechanics and Thermodynamics
(HEFAT), pp. 6, Sun City, South Africa, 2007.
[36] Baker, D. K., and Kaftanolu, B., "Gne Enerjisi Ile alan Adsorplanma
Soutma Sisteminden Maksimum Teorik Baarm Salanmas," Proceedings of
VI. Ulusal Temiz Enerji Sempozyumu (UTES), Isparta, Turkey, 2006.
118
[37] Qu, T. F., Wang, R. Z., and Wang, W., "Study on Heat and Mass Recovery in
Adsorption Refrigeration Cycles," Applied Thermal Engineering, vol. 21, no.
4, pp. 439-452, 2001.
[38] Wang, R. Z., "Performance Improvement of Adsorption Cooling by Heat and
Mass Recovery Operation," International Journal of Refrigeration, vol. 24, no.
7, pp. 602-611, 2001.
[39] Wang, X., and Chua, H. T., "A Comparative Evaluation of Two Different
Heat-Recovery Schemes as Applied to a Two-Bed Adsorption Chiller,"
International Journal of Heat and Mass Transfer, vol. 50, no. 3-4, pp. 433-443,
2007.
[40] Akahira, A., Alam, K. C. A., Hamamoto, Y., Akisawa, A., and Kashiwagi, T.,
"Mass Recovery Adsorption Refrigeration Cycle - Improving Cooling
Capacity," International Journal of Refrigeration, vol. 27, no. 3, pp. 225-234,
2004.
[41] Wang, W., Qu, T. F., and Wang, R. Z., "Influence of Degree of Mass Recovery
and Heat Regeneration on Adsorption Refrigeration Cycles," Energy
Conversion and Management, vol. 43, no. 5, pp. 733-741, 2002.
[42] Shelton, S. V., Wepfer, W. J., and Miles, D. J., "Ramp Wave Analysis of the
Solid/Vapor Heat Pump," Journal of Energy Resources Technology, vol. 112,
no. 1, pp. 69-78, 1990.
[43] Shelton, S. V., "Solid Adsorbent Heat Pump System," US Patent 4610148,
USA, 1986.
[44] Taylan, O., Baker, D. K., and Kaftanolu, B., "COP Trends for Ideal Thermal
Wave Adsorption Cooling Cycles with Enhancements," International Journal
of Refrigeration: Under Review, 2009.
[45] Critoph, R. E., "Forced Convection Adsorption Cycles," Applied Thermal
Engineering, vol. 18, no. 9-10, pp. 799-807, 1998.
[46] Sanborn, D. M., "Direct Heated Adsorbent Bed Heat Pump," US Patent
5505059, USA, 1996.
[47] Maier-Laxhuber, P., and Kaubek, F., "Continuously Acting Adsorption
Devices and Process for Their Operation," US Patent 4660629, USA, 1987.
[48] Erickson, D. C., "Rotary Trisorption Heat Pump," US Patent 5279359, USA,
1994.
[49] Ebbeson, B., "Sorption Device," US Patent 5431716, USA, 1995.
119
[50] Llobet, J., and Goetz, V., "Rotary System for the Continuous Production of
Cold by Solid-Gas Sorption: Modeling and Analysis of Energy Performance,"
International Journal of Refrigeration, vol. 23, no. 8, pp. 609-625, 2000.
[51] Chua, H. T., Ng, K. C., Malek, A., Kashiwagi, T., Akisawa, A., and Saha, B.
B., "Multi-Bed Regenerative Adsorption Chiller - Improving the Utilization of
Waste Heat and Reducing the Chilled Water Outlet Temperature Fluctuation,"
[52] Yong, L., and Sumathy, K., "Performance Analysis of a Continuous Multi-Bed
Adsorption Rotary Cooling System," Applied Thermal Engineering, vol. 25,
no. 2-3, pp. 393-407, 2005.
[53] De Boer, R., and Smeding, S. F., "Development and Testing of a Rotating
System for a Continuous Solid-Sorption Process," Proceedings of Heat
Powered Cycles Conference, Newcastle upon Tyne, 2006.
[54] Pons, M., and Guilleminot, J. J., "Design of an Experimental Solar-Powered,
Solid-Adsorption Ice Maker," Journal of Solar Energy Engineering, vol. 108,
no. 4, pp. 332-337, 1986.
[55] Wang, W., Wang, R. Z., Xu, Y. X., Wu, J. Y., and Gui, Y. B., "Investigation
on Adsorption Refrigeration with a Single Adsorbent Bed," International
Journal of Energy Research, vol. 22, no. 13, pp. 1157-1163, 1998.
[56] Pons, M., Meunier, F., Cacciola, G., Critoph, R. E., Groll, M., Puigjaner, L.,
Spinner, B., and Ziegler, F., "Thermodynamic Based Comparison of Sorption
Systems for Cooling and Heat Pumping," International Journal of
Refrigeration, vol. 22, pp. 5-17, 1999.
[57] Critoph, R., Tamainot-Telto, Z., and Davies, G., "A Prototype of a Fast Cycle
Adsorption Refrigerator Utilizing a Novel Carbon - Aluminium Laminate,"
Proceedings of the Institution of Mechanical Engineers, Part A: Journal of
Power and Energy, vol. 214, no. 5, pp. 439-448, 2000.
[58] Saha, B. B., Akisawa, A., and Kashiwagi, T., "Solar/Waste Heat Driven Two-
Stage Adsorption Chiller: The Prototype," Renewable Energy, vol. 23, no. 1,
pp. 93-101, 2001.
[59] Anyanwu, E. E., and Ezekwe, C. I., "Design, Construction and Test Run of a
Solid Adsorption Solar Refrigerator Using Activated Carbon/Methanol, as
Adsorbent/Adsorbate Pair," Energy Conversion and Management, vol. 44, no.
18, pp. 2879-2892, 2003.
[60] Wang, L. W., Wu, J. Y., Wang, R. Z., Xu, Y. X., and Wang, S. G.,
"Experimental Study of a Solidified Activated Carbon-Methanol Adsorption
Ice Maker," Applied Thermal Engineering, vol. 23, no. 12, pp. 1453-1462,
2003.
120
[61] Wang, L. W., Wu, J. Y., Wang, R. Z., Xu, Y. X., Wang, S. G., and Li, X. R.,
"Study of the Performance of Activated Carbon-Methanol Adsorption Systems
Concerning Heat and Mass Transfer," Applied Thermal Engineering, vol. 23,
no. 13, pp. 1605-1617, 2003.
[62] Wang, S., and Zhu, D., "Adsorption Heat Pump Using an Innovative Coupling
Refrigeration Cycle," Adsorption, vol. 10, no. 1, pp. 47-55, 2004.
[63] Chang, W. S., Wang, C. C., and Shieh, C. C., "Experimental Study of a Solid
Adsorption Cooling System Using Flat-Tube Heat Exchangers as Adsorption
Bed," Applied Thermal Engineering, vol. 27, no. 13, pp. 2195-2199, 2007.
[64] Zhai, X. Q., and Wang, R. Z., "Experimental Investigation and Theoretical
Analysis of the Solar Adsorption Cooling System in a Green Building,"
Applied Thermal Engineering, vol. 29, no. 1, pp. 17-27, 2009.
[65] Banker, N. D., Prasad, M., Dutta, P., and Srinivasan, K., "Development and
Transient Performance Results of a Single Stage Activated Carbon - HFC 134a
Closed Cycle Adsorption Cooling System," Applied Thermal Engineering: In
Press, Corrected Proof, 2010.
[66] Grisel, R. J. H., Smeding, S. F., and Boer, R. D., "Waste Heat Driven Silica
Gel/Water Adsorption Cooling in Trigeneration," Applied Thermal
Engineering, vol. 30, no. 8-9, pp. 1039-1046, 2010.
[67] Baker, D. K., and Kaftanolu, B., "Trends in COP for Adsorption Cooling
Cycles with Thermal Regeneration and Finite Number of Beds," Proceedings
of Energy Sustainability 2008, Jacksonville, Florida, USA, 2008.
[68] Myers, A. L., "Thermodynamics of Adsorption in Porous Materials," AIChE
Journal, vol. 48, no. 1, pp. 145-160, 2002.
[69] Pons, M., and Kodama, A., "Entropic Analysis of Adsorption Open Cycles for
Air Conditioning. Part 1: First and Second Law Analyses," International
[70] Douss, N., Meunier, F. E., and Sun, L.-M., "Predictive Model and
Experimental Results for a Two-Adsorber Solid Adsorption Heat Pump,"
Industrial & Engineering Chemistry Research, vol. 27, no. 2, 1988.
[71] Zheng, W., Worek, W. M., and Nowakowski, G., "Effect of Operating
Conditions on the Performance of Two-Bed Closed-Cycle Solid-Sorption Heat
Pump Systems," Journal of Solar Energy Engineering, vol. 117, no. 3, pp. 181-
186, 1995.
[72] Zheng, W., Worek, W. M., and Nowakowski, G., "Effect of Design and
Operating Parameters on the Performance of Two-Bed Sorption Heat Pump
Systems," Journal of Energy Resources Technology, vol. 117, no. 1, pp. 67-74,
1995.
121
[73] Teng, Y., Wang, R. Z., and Wu, J. Y., "Study of the Fundamentals of
Adsorption Systems," Applied Thermal Engineering, vol. 17, pp. 327-338,
1997.
[74] Sun, L. M., Feng, Y., and Pons, M., "Numerical Investigation of Adsorptive
Heat Pump Systems with Thermal Wave Heat Regeneration under Uniform-
Pressure Conditions," International Journal of Heat and Mass Transfer, vol.
40, no. 2, pp. 281-293, 1997.
[75] Pons, M., and Poyelle, F., "Adsorptive Machines with Advanced Cycles for
Heat Pumping or Cooling Applications," International Journal of
Refrigeration, vol. 22, no. 1, pp. 27-37, 1999.
[76] Sward, B. K., Levan, M. D., and Meunier, F., "Adsorption Heat Pump
Modeling: The Thermal Wave Process with Local Equilibrium," Applied
Thermal Engineering, vol. 20, no. 8, pp. 759-780, 2000.
[77] Wang, R. Z., Li, M., Xu, Y. X., and Wu, J. Y., "An Energy Efficient Hybrid
System of Solar Powered Water Heater and Adsorption Ice Maker," Solar
Energy, vol. 68, no. 2, pp. 189-195, 2000.
[78] Pons, M., and Szarzynski, S., "Accounting for the Real Properties of the Heat
Transfer Fluid in Heat-Regenerative Adsorption Cycles for Refrigeration,"
[79] Saha, B. B., Koyama, S., Lee, J. B., Kuwahara, K., Alam, K. C. A., Hamamoto,
Y., Akisawa, A., and Kashiwagi, T., "Performance Evaluation of a Low-
Temperature Waste Heat Driven Multi-Bed Adsorption Chiller," International
Journal of Multiphase Flow, vol. 29, no. 8, pp. 1249-1263, 2003.
[80] Alam, K. C. A., Saha, B., Akisawa, A., and Kashiwagi, T., "Influence of
Design and Operating Conditions on the System Performance of a Two-Stage
Adsorption Chiller," Chemical Engineering Communications, vol. 191, pp.
981-997, 2004.
[81] Wang, D. C., Xia, Z. Z., Wu, J. Y., Wang, R. Z., Zhai, H., and Dou, W. D.,
"Study of a Novel Silica Gel-Water Adsorption Chiller. Part I. Design and
Performance Prediction," International Journal of Refrigeration, vol. 28, no. 7,
pp. 1073-1083, 2005.
[82] Lambert, M. A., "Design of Solar Powered Adsorption Heat Pump with Ice
Storage," Applied Thermal Engineering, vol. 27, no. 8-9, pp. 1612-1628, 2007.
[83] Liu, Y., and Leong, K. C., "Numerical Modeling of a Zeolite/Water Adsorption
Cooling System with Non-Constant Condensing Pressure," International
Communications in Heat and Mass Transfer, vol. 35, no. 5, pp. 618-622, 2008.
122
[84] Wang, D. C., and Zhang, J. P., "Design and Performance Prediction of an
Adsorption Heat Pump with Multi-Cooling Tubes," Energy Conversion and
Management, vol. 50, no. 5, pp. 1157-1162, 2009.
[85] Luo, H., Wang, R. Z., and Dai, Y. J., "Optimum Matching of Heat Source
Temperature to a Solar Adsorption Air-Conditioning System for Maximum
Solar Cooling Coefficient of Performance," International Journal of Green
Energy, vol. 7, no. 1, pp. 91 - 102, 2010.
[86] Aristov, Y. I., "Novel Materials for Adsorptive Heat Pumping and Storage:
Screening and Nanotailoring of Sorption Properties," Journal of Chemical
Engineering of Japan, vol. 40, no. 13, pp. 1242-1251, 2007.
[87] Solmu, ., Yamal, C., Kaftanolu, B., Baker, D. K., and alar, A.,
"Adsorption Properties of a Natural Zeolite-Water Pair for Use in Adsorption
Cooling Cycles," Applied Energy, In Press, Corrected Proof, 2009.
[88] Wang, R. Z., and Wang, Q. B., "Adsorption Mechanism and Improvements of
the Adsorption Equation for Adsorption Refrigeration Pairs," International
[89] Taylan, O., Baker, D. K., and Kaftanolu, B., "Adsorbent - Refrigerant
Comparison for a Solar Powered Adsorption Cooling System Using Seasonal
Simulations," Proceedings of 10th REHVA World Congress, Antalya, Turkey,
2010.
[90] Baker, D. K., "Thermodynamic Limits to Thermal Regeneration in Adsorption
Cooling Cycles," International Journal of Refrigeration, vol. 31, no. 1, pp. 55-
64, 2008.
[91] Lu, Y. Z., Wang, R. Z., Zhang, M., and Jiangzhou, S., "Adsorption Cold
Storage System with Zeolite-Water Working Pair Used for Locomotive Air
Conditioning," Energy Conversion and Management, vol. 44, no. 10, pp. 1733-
1743, 2003.
[92] Daou, K., Wang, R. Z., Yang, G. Z., and Xia, Z. Z., "Theoretical Comparison
of the Refrigerating Performances of a CaCl
2
Impregnated Composite
Adsorbent to Those of the Host Silica Gel," International Journal of Thermal
Sciences, vol. 47, no. 1, pp. 68-75, 2008.
[93] Wang, L. W., Wang, R. Z., Lu, Z. S., Chen, C. J., Wang, K., and Wu, J. Y.,
"The Performance of Two Adsorption Ice Making Test Units Using Activated
Carbon and a Carbon Composite as Adsorbents," Carbon, vol. 44, no. 13, pp.
2671-2680, 2006.
[94] Ben Amar, N., Sun, L. M., and Meunier, F., "Numerical Analysis of
Adsorptive Temperature Wave Regenerative Heat Pump " Applied Thermal
Engineering, vol. 16, pp. 405-418, 1996.
123
[95] Saha, B. B., Habib, K., El-Sharkawy, I. I., and Koyama, S., "Adsorption
Characteristics and Heat of Adsorption Measurements of R-134a on Activated
Carbon," International Journal of Refrigeration, vol. 32, no. 7, pp. 1563-1569,
2009.
[96] Wang, R. Z., Jia, J. P., Zhu, Y. H., Teng, Y., Wu, J. Y., Cheng, J., and Wang,
Q. B., "Study on a New Solid Adsorption Refrigeration Pair: Active Carbon
Fiber-Methanol," Journal of Solar Energy Engineering, vol. 119, no. 3, pp.
214-218, 1997.
[97] Restuccia, G., Freni, A., Vasta, S., and Aristov, Y., "Selective Water Sorbent
for Solid Sorption Chiller: Experimental Results and Modelling," International
Journal of Refrigeration, vol. 27, no. 3, pp. 284-293, 2004.
[98] Wang, L. W., Wang, R. Z., Wu, J. Y., and Wang, K., "Compound Adsorbent
for Adsorption Ice Maker on Fishing Boats," International Journal of
[99] Cui, Q., Tao, G., Chen, H., Guo, X., and Yao, H., "Environmentally Benign
Working Pairs for Adsorption Refrigeration," Energy, vol. 30, no. 2-4, pp. 261-
271, 2005.
[100] Hamamoto, Y., Alam, K. C. A., Saha, B. B., Koyama, S., Akisawa, A., and
Kashiwagi, T., "Study on Adsorption Refrigeration Cycle Utilizing Activated
Carbon Fibers. Part 2. Cycle Performance Evaluation," International Journal of
[101] Saha, B. B., El-Sharkawy, I. I., Chakraborty, A., Koyama, S., Banker, N. D.,
Dutta, P., Prasad, M., and Srinivasan, K., "Evaluation of Minimum Desorption
Temperatures of Thermal Compressors in Adsorption Refrigeration Cycles,"
[102] El-Sharkawy, I. I., Saha, B. B., Koyama, S., He, J., Ng, K. C., and Yap, C.,
"Experimental Investigation on Activated Carbon-Ethanol Pair for Solar
Powered Adsorption Cooling Applications," International Journal of
[103] Loh, W. S., El-Sharkawy, I. I., Ng, K. C., and Saha, B. B., "Adsorption
Cooling Cycles for Alternative Adsorbent/Adsorbate Pairs Working at Partial
Vacuum and Pressurized Conditions," Applied Thermal Engineering, vol. 29,
no. 4, pp. 793-798, 2009.
[104] "TRNSYS," The University of Wisconsin-Madison Solar Energy Lab,
Madison, 2007.
[105] TRNSYS: The Transient Energy System Simulation Tool, Last accessed March
16, 2010, http://www.trnsys.com/.
124
[106] TRNSYS Overview, Last accessed March 16, 2010, http://www.tess-
inc.com/trnsys.
[107] A Transient Systems Simulation Program, Last accessed March 16, 2010,
http://sel.me.wisc.edu/trnsys/default.htm.
[108] Taylan, O., Baker, D. K., and Kaftanolu, B., "Parametric Study and Seasonal
Simulations of a Solar Powered Adsorption Cooling System," Proceedings of
22nd International Conference on Efficiency, Cost, Optimization, Simulation
and Environmental Impact of Energy Systems (ECOS), pp. 833 - 842, Foz do
Iguau, Paran, Brazil, 2009.
[109] "Methodology for Determining the Thermal Performance Rating for Solar
Collectors," Solar Rating and Certification Corporation, Cocoa, 1994.
[110] Schmidt, P. S., Ezekoye, O., Howell, J. R., and Baker, D. K., Thermodynamics:
An Integrated Learning System, Wiley, 2005.
[111] Kaftanolu, B., Yamal, C., Baker, D. K., Solmu, ., alar, A., and Taylan,
O., "Solar Powered Zeolite Air-Conditioning System," The Scientific and
Technological Council of Turkey (TBTAK) Project No: 105M244, Ankara,
2009.
[112] Kalogirou, S. A., "Solar Thermal Collectors and Applications," Progress in
Energy and Combustion Science, vol. 30, no. 3, pp. 231-295, 2004.
[113] Lin, G., and Yan, Z., "The Optimal Operating Temperature of the Collector of
an Irreversible Solar-Driven Refrigerator," Journal of Physics D: Applied
Physics, vol. 32, no. 2, pp. 94, 1999.
[114] alar, A., Yamal, C., Baker, D. K., and Kaftanolu, B., "Testing the Thermal
Performances of Solar Collectors " Proceedings of 7
th
International Conference
on Heat Transfer, Fluid Mechanics and Thermodynamics (HEFAT 2010),
Antalya, Turkey, 2010.

125
APPENDIX A

SAMPLE TRNSYS INPUT (DECK) FILE
VERSION 16.1
********************************************************************
*** TRNSYS input file (deck) generated by TrnsysStudio
***
*** If you edit this file, use the File/Import TRNSYS Input File function in
*** TrnsysStudio to update the project.
***
*** If you have problems, questions or suggestions please contact your local
*** TRNSYS distributor or mailto:software@cstb.fr
***
********************************************************************

********************************************************************
*** Units
********************************************************************

********************************************************************
*** Control cards
********************************************************************
* START, STOP and STEP
CONSTANTS 3
START=3624
STOP=6552
STEP=0.25
126
* User defined CONSTANTS
SIMULATION START STOP STEP ! Start time End time
Time step
TOLERANCES 0.000001 0.000001 ! Integration Conver-
gence
LIMITS 30 30 30 ! Max iterations Max warnings Trace limit
DFQ 1 ! TRNSYS numerical integration solver method
WIDTH 80 ! TRNSYS output file width, number of charac-
ters
LIST ! NOLIST statement
! MAP statement
SOLVER 0 1 1 ! Solver statement Minimum relaxation factor
Maximum relaxation factor
NAN_CHECK 0 ! Nan DEBUG statement
OVERWRITE_CHECK 0 ! Overwrite DEBUG statement
TIME_REPORT 0 ! disable time report
EQSOLVER 0 ! EQUATION SOLVER statement

* EQUATIONS "Equa"
*
EQUATIONS 6
AdsNo = 1
Tref = 21
Tcold = 10
Tadsmax = 90
DTexcess = 0
R = 10
*$UNIT_NAME Equa
*$LAYER Main
*$POSITION 212 125
*------------------------------------------------------------------------------
* Model "Type155" (Type 155)
127
*
UNIT 11 TYPE 155 Type155
*$UNIT_NAME Type155
*$MODEL .\Utility\Calling External Programs\Matlab\Type155.tmf
*$POSITION 105 479
*$LAYER Main #
PARAMETERS 5
0 ! 1 Mode
7 ! 2 Number of inputs
6 ! 3 Number of outputs
0 ! 4 Calling Mode
0 ! 5 Keep Matlab open after simulation
INPUTS 7
16,1 ! Type92:Outlet fluid temperature ->input-1
Tbedmin ! Equa-2:Tbedmin ->input-2
17,1 ! Type109-TMY2:Ambient temperature ->input-3
Tcold ! Equa:Tcold ->input-4
R ! Equa:R ->input-5
AdsNo ! Equa:AdsNo ->input-6
14,1 ! Type2b:Output control function ->input-7
*** INITIAL INPUT VALUES
90 30 30 10 0 2 0
LABELS 1
" \..\..\..\..\sample.m"
*------------------------------------------------------------------------------
* Model "Type25c" (Type 25)
*
UNIT 9 TYPE 25 Type25c
*$UNIT_NAME Type25c
*$MODEL .\Output\Printer\Unformatted\No Units\Type25c.tmf
*$POSITION 692 125
*$LAYER Outputs #
128
PARAMETERS 10
STEP ! 1 Printing interval
START ! 2 Start time
STOP ! 3 Stop time
35 ! 4 Logical unit
0 ! 5 Units printing mode
0 ! 6 Relative or absolute start time
1 ! 7 Overwrite or Append
-1 ! 8 Print header
0 ! 9 Delimiter
1 ! 10 Print labels
INPUTS 16
hColl ! Equa-2:hColl ->Input to be printed-1
11,1 ! Type155:output-1 ->Input to be printed-2
COPsys_rev ! Equa-2:COPsys_rev ->Input to be printed-5
COPsystww ! Equa-2:COPsys_tw_w ->Input to be printed-6
COPsystwwo ! Equa-2:COPsys_tw_wo ->Input to be printed-7
Qload ! Equa-2:Qload ->Input to be printed-8
19,3 ! Type71:Useful energy gain ->Input to be printed-9
16,5 ! Type92:Rate of energy removed ->Input to be printed-10
Qclg_rev ! Equa-2:Qclg_rev ->Input to be printed-11
Qclg_tw_w ! Equa-2:Qclg_tw_w ->Input to be printed-12
Qclg_tw_wo ! Equa-2:Qclg_tw_wo ->Input to be printed-13
DX_DXref ! Equa-2:DX_DXref ->Input to be printed-14
hColl COPrev COPw COPwo COPsys_rev COPsys_w COPsys_wo Qload
Qcoll Qaux
Qclg_rev Qclg_w Qclg_wo m_ads Errw Errwo
129
*** External files
ASSIGN "Ad_Z1_90.out" 35
*|? Output file for printed results |1000
*------------------------------------------------------------------------------
* Model "Type65d" (Type 65)
*
UNIT 15 TYPE 65 Type65d
*$UNIT_NAME Type65d
*$MODEL .\Output\Online Plotter\Online Plotter Without File\Type65d.tmf
*$POSITION 687 378
*$LAYER Main #
PARAMETERS 12
3 ! 1 Nb. of left-axis variables
2 ! 2 Nb. of right-axis variables
-2 ! 3 Left axis minimum
2 ! 4 Left axis maximum
0.0 ! 5 Right axis minimum
250 ! 6 Right axis maximum
1 ! 7 Number of plots per simulation
12 ! 8 X-axis gridpoints
0 ! 9 Shut off Online w/o removing
-1 ! 10 Logical unit for output file
0 ! 11 Output file units
0 ! 12 Output file delimiter
INPUTS 5
11,1 ! Type155:output-1 ->Left axis variable-1
hColl ! Equa-2:hColl ->Right axis variable-1
16,1 ! Type92:Outlet fluid temperature ->Right axis variable-2
COPrev COPnobypass COPbypass hColl Texit
130
LABELS 3
"Efficiencies"
""
"Graph 1"
*------------------------------------------------------------------------------
* EQUATIONS "Equa-2"
*
EQUATIONS 10
Tbedmin = [17,1]+DTexcess
Qload = [17,1]-Tref
hColl = [19,3]/1.62/([17,18]+1e-16)*gt([17,18],0)
DX_DXref = [11,4]/0.125794687663051
Qclg_rev = [16,5]*[11,1]
Qclg_tw_w = [16,5]*[11,2]
Qclg_tw_wo = [16,5]*[11,3]
COPsys_rev = hColl*[11,1]
COPsys_tw_w = hColl*[11,2]
COPsys_tw_wo = hColl*[11,3]
*$UNIT_NAME Equa-2
*$LAYER Main
*$POSITION 337 370
*------------------------------------------------------------------------------
* Model "Type2b" (Type 2)
*
UNIT 14 TYPE 2 Type2b
*$UNIT_NAME Type2b
*$MODEL .\Controllers\Differential Controller w_ Hysteresis\for Tempera-
tures\Solver 0 (Successive Substitution) Control Strategy\Type2b.tmf
*$POSITION 467 498
*$LAYER Controls #
*$# NOTE: This control strategy can only be used with solver 0 (Successive
substitution)
131
*$#
PARAMETERS 2
5 ! 1 No. of oscillations
100.0 ! 2 High limit cut-out
INPUTS 6
19,1 ! Type71:Outlet temperature ->Upper input temperature Th
Tadsmax ! Equa:Tadsmax ->Lower input temperature Tl
0,0 ! [unconnected] Monitoring temperature Tin
0,0 ! [unconnected] Input control function
0,0 ! [unconnected] Upper dead band dT
0,0 ! [unconnected] Lower dead band dT
100 150 20.0 0 0 0
*------------------------------------------------------------------------------
*
*$UNIT_NAME Type92
*$MODEL .\HVAC\Auxiliary Cooling Unit\Type92.tmf
*$POSITION 646 294
*$LAYER Water Loop #
*$# This model is extremely similar to the TYPE6 Auxiliary Heater
PARAMETERS 4
100000 ! 1 Maximum cooling rate
4.19 ! 2 Specific heat of fluid
0 ! 3 Overall loss coefficient
1.0 ! 4 Cooling device efficiency
INPUTS 5
19,1 ! Type71:Outlet temperature ->Inlet fluid temperature
19,2 ! Type71:Outlet flowrate ->Fluid mass flow rate
14,1 ! Type2b:Output control function ->Control Function
Tadsmax ! Equa:Tadsmax ->Set point temperature
132
17,1 ! Type109-TMY2:Ambient temperature ->Temperature of sur-
roundings
20 100 1 150 20
*------------------------------------------------------------------------------
* Model "Type109-TMY2" (Type 109)
*
UNIT 17 TYPE 109 Type109-TMY2
*$UNIT_NAME Type109-TMY2
*$MODEL .\Weather Data Reading and Processing\Standard For-
mat\TMY2\Type109-TMY2.tmf
*$POSITION 287 157
*$LAYER Weather - Data Files #
PARAMETERS 4
2 ! 1 Data Reader Mode
41 ! 2 Logical unit
4 ! 3 Sky model for diffuse radiation
1 ! 4 Tracking mode
INPUTS 3
0,0 ! [unconnected] Ground reflectance
0,0 ! [unconnected] Slope of surface
0,0 ! [unconnected] Azimuth of surface
0.2 40 0.0
*** External files
ASSIGN "\..\..\..\..\Antalya.tm2" 41
*|? Weather data file |1000
*------------------------------------------------------------------------------
* Model "Type3d" (Type 3)
*
UNIT 18 TYPE 3 Type3d
*$UNIT_NAME Type3d
133
*$MODEL .\Hydronics\Pumps\Single Speed - no Powercoeffi-
cients\Type3d.tmf
*$POSITION 53 402
*$LAYER Main #
PARAMETERS 4
100 ! 1 Maximum flow rate
4.19 ! 2 Fluid specific heat
60 ! 3 Maximum power
0 ! 4 Conversion coefficient
INPUTS 3
16,1 ! Type92:Outlet fluid temperature ->Inlet fluid temperature
16,2 ! Type92:Outlet fluid flow rate ->Inlet mass flow rate
0,0 ! [unconnected] Control signal
20 100 1
*------------------------------------------------------------------------------
*
*$UNIT_NAME Type71
*$MODEL .\Solar Thermal Collectors\Evacuated Tube Collector\Type71.tmf
*$POSITION 287 294
*$LAYER Main #
PARAMETERS 11
1 ! 1 Number in series
2 ! 2 Collector area
4.19 ! 3 Fluid specific heat
2 ! 4 Efficiency mode
3.0 ! 5 Flow rate at test conditions
0.825 ! 6 Intercept efficiency
4.284 ! 7 Negative of first order efficiency coeficient
0.0324 ! 8 Negative of second order efficiency coeficient
134
42 ! 9 Logical unit of file containing biaxial IAM data
10 ! 10 Number of longitudinal angles for which IAMs are pro-
vided
10 ! 11 Number of transverse angles for which IAMs are provid-
ed
INPUTS 10
18,1 ! Type3d:Outlet fluid temperature ->Inlet temperature
18,2 ! Type3d:Outlet flow rate ->Inlet flowrate
17,1 ! Type109-TMY2:Ambient temperature ->Ambient tempera-
ture
17,18 ! Type109-TMY2:total radiation on tilted surface ->Incident
radiation
17,20 ! Type109-TMY2:sky diffuse radiation on tilted surface -
>Incident diffuse radiation
17,22 ! Type109-TMY2:angle of incidence for tilted surface ->Solar
incidence angle
17,10 ! Type109-TMY2:solar zenith angle ->Solar zenith angle
17,11 ! Type109-TMY2:solar azimuth angle ->Solar azimuth angle
17,23 ! Type109-TMY2:slope of tilted surface ->Collector slope
0,0 ! [unconnected] Collector azimuth
20 100 10 0 0 0.0 0.0 0.0 40 0.0
*** External files
ASSIGN "..\..\..\Tess Models\SampleCatalogData\Bi-Axial IAMs\Iam_2d.dat"
42
*|? What file contains the 2D IAM data? |1000
*------------------------------------------------------------------------------
END
135
APPENDIX B

SAMPLE TRNSYS STUDIO REPORT
********************************************************************
*** Control cards
********************************************************************
TOLERANCES Integration 0.000001 Convergence 0.000001
LIMITS
Max iterations: 30
Max warnings: 30
Trace limit: 30
TRNSYS numerical integration solver method: DFQ 1
TRNSYS output file width, number of characters:
WIDTH 80
NOLIST statement: LIST
MAP statement: none
Solver statement: SOLVER 0
START=3624
STOP=6552
STEP=0.25
Write TRNSED commands : OFF
* User defined CONSTANTS
********************************************************************
*** Simulation order
********************************************************************
1. Model: CONTROL CARD
2. Model: Equation Editor (Equa)
3. Model: MATLAB Caller (Type155)
136
4. Model: Output to File (Type25c)
5. Model: Output to Screen (Type65d)
6. Model: Equation Editor (Equa-2)
7. Macro: Solar Thermal System (Macro)
8. Model: Control Unit (Type2b)
9. Model: Heat Exchanger (Type92)
10. Model: Weather Data Reader (Type109-TMY2)
11. Model: Circulation Pump (Type3d)
12. Model: Evacuated Tube Collector (Type71)
********************************************************************
*** Macro structure
********************************************************************
Output to File (Type25c)
MATLAB Caller (Type155)
Equation Editor (Equa)
Output to Screen (Type65d)
Equation Editor (Equa-2)
Solar Thermal System (Macro)
Control Unit (Type2b)
Heat Exchanger (Type92)
Weather Data Reader (Type109-TMY2)
Circulation Pump (Type3d)
Evacuated Tube Collector (Type71)
********************************************************************
*** Unconnected inputs
********************************************************************
CONTROL CARD (Unit 1, Type 0)
------------------------------------------------------------------------------
MATLAB Caller (Type155) (Unit 3, Type 155)
------------------------------------------------------------------------------
Output to File (Type25c) (Unit 4, Type 25)
------------------------------------------------------------------------------
137
Output to Screen (Type65d) (Unit 5, Type 65)
------------------------------------------------------------------------------
Control Unit (Type2b) (Unit 7, Type 2)
Monitoring temperature Tin - constant value is 20.0 [C]
Input control function - constant value is 0 [-]
Upper dead band dT - constant value is 0 [Temp. Difference]
Lower dead band dT - constant value is 0 [Temp. Difference]
------------------------------------------------------------------------------
Heat Exchanger (Type92) (Unit 8, Type 92)
------------------------------------------------------------------------------
Weather Data Reader (Type109-TMY2) (Unit 9, Type 109)
Ground reflectance - constant value is 0.2 [-]
Slope of surface - constant value is 40 [degrees]
Azimuth of surface - constant value is 0.0 [degrees]
------------------------------------------------------------------------------
Circulation Pump (Type3d) (Unit 10, Type 3)
Control signal - constant value is 1 [-]
------------------------------------------------------------------------------
Evacuated Tube Collector (Type71) (Unit 11, Type 71)
Collector azimuth - constant value is 0.0 [degrees]
------------------------------------------------------------------------------
********************************************************************
*** All connections
********************************************************************
CONTROL CARD (Unit 1, Type 0)
------------------------------------------------------------------------------
MATLAB Caller (Type155) (Unit 3, Type 155)
output-1 -> Left axis variable-1 (Output to Screen (Type65d))
output-1 -> COPrev (Equation Editor (Equa-2))
output-1 -> Input to be printed-2 (Output to File (Type25c))
output-2 -> COPtw_w (Equation Editor (Equa-2))
138
output-3 -> COPtw_wo (Equation Editor (Equa-2))
output-4 -> DX (Equation Editor (Equa-2))
------------------------------------------------------------------------------
Output to File (Type25c) (Unit 4, Type 25)
------------------------------------------------------------------------------
Output to Screen (Type65d) (Unit 5, Type 65)
------------------------------------------------------------------------------
Control Unit (Type2b) (Unit 7, Type 2)
Output control function -> Control Function (Heat Exchanger (Type92))
Output control function -> input-7 (MATLAB Caller (Type155))
------------------------------------------------------------------------------
Heat Exchanger (Type92) (Unit 8, Type 92)
Outlet fluid temperature -> Inlet fluid temperature (Circulation Pump
(Type3d))
Outlet fluid temperature -> input-1 (MATLAB Caller (Type155))
Outlet fluid temperature -> Right axis variable-2 (Output to Screen
(Type65d))
Outlet fluid flow rate -> Inlet mass flow rate (Circulation Pump (Type3d))
Rate of energy removed -> Qaux (Equation Editor (Equa-2))
Rate of energy removed -> Input to be printed-10 (Output to File (Type25c))
------------------------------------------------------------------------------
Weather Data Reader (Type109-TMY2) (Unit 9, Type 109)
Ambient temperature -> Temperature of surroundings (Heat Exchanger
(Type92))
Ambient temperature -> Ambient temperature (Evacuated Tube Collector
(Type71))
Ambient temperature -> Tamb (Equation Editor (Equa-2))
139
Ambient temperature -> input-3 (MATLAB Caller (Type155))
Ambient temperature -> Tcond (Equation Editor (Equa-2))
solar zenith angle -> Solar zenith angle (Evacuated Tube Collector (Type71))
solar azimuth angle -> Solar azimuth angle (Evacuated Tube Collector
(Type71))
total radiation on tilted surface -> Incident radiation (Evacuated Tube Collec-
tor (Type71))
total radiation on tilted surface -> Radiation (Equation Editor (Equa-2))
sky diffuse radiation on tilted surface -> Incident diffuse radiation (Evacuated
Tube Collector (Type71))
angle of incidence for tilted surface -> Solar incidence angle (Evacuated
Tube Collector (Type71))
slope of tilted surface -> Collector slope (Evacuated Tube Collector
(Type71))
------------------------------------------------------------------------------
Circulation Pump (Type3d) (Unit 10, Type 3)
Outlet fluid temperature -> Inlet temperature (Evacuated Tube Collector
(Type71))
Outlet flow rate -> Inlet flowrate (Evacuated Tube Collector (Type71))
------------------------------------------------------------------------------
Evacuated Tube Collector (Type71) (Unit 11, Type 71)
Outlet temperature -> Upper input temperature Th (Control Unit (Type2b))
Outlet temperature -> Inlet fluid temperature (Heat Exchanger (Type92))
Outlet flowrate -> Fluid mass flow rate (Heat Exchanger (Type92))
Useful energy gain -> Qcoll (Equation Editor (Equa-2))
Useful energy gain -> Input to be printed-9 (Output to File (Type25c))
------------------------------------------------------------------------------
End of report.

Index

Hochgeladen von

Dokumentinformationen

Originalbeschreibung:

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Index

Hochgeladen von

Copyright:

Verfügbare Formate

NUMERICAL MODELING AND PERFORMANCE ANALYSIS

OF SOLAR-POWERED IDEAL ADSORPTION COOLING SYSTEMS

mean amb mean amb

Das könnte Ihnen auch gefallen